8th day of python challenges 111-117

This commit is contained in:
abd.shallal
2019-08-04 15:26:35 +03:00
parent b04c1b055f
commit 627802c383
3215 changed files with 760227 additions and 491 deletions

View File

@@ -0,0 +1,134 @@
# Copyright (c) 2006-2014 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2013-2014 Google, Inc.
# Copyright (c) 2013 buck@yelp.com <buck@yelp.com>
# Copyright (c) 2014-2017 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Brett Cannon <brett@python.org>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Moises Lopez <moylop260@vauxoo.com>
# Copyright (c) 2017-2018 Bryce Guinta <bryce.paul.guinta@gmail.com>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""utilities methods and classes for checkers
Base id of standard checkers (used in msg and report ids):
01: base
02: classes
03: format
04: import
05: misc
06: variables
07: exceptions
08: similar
09: design_analysis
10: newstyle
11: typecheck
12: logging
13: string_format
14: string_constant
15: stdlib
16: python3
17: refactoring
18-50: not yet used: reserved for future internal checkers.
51-99: perhaps used: reserved for external checkers
The raw_metrics checker has no number associated since it doesn't emit any
messages nor reports. XXX not true, emit a 07 report !
"""
import sys
import tokenize
import warnings
from typing import Any
from pylint.config import OptionsProviderMixIn
from pylint.reporters import diff_string
from pylint.utils import register_plugins
from pylint.interfaces import UNDEFINED
def table_lines_from_stats(stats, old_stats, columns):
"""get values listed in <columns> from <stats> and <old_stats>,
and return a formated list of values, designed to be given to a
ureport.Table object
"""
lines = []
for m_type in columns:
new = stats[m_type]
format = str # pylint: disable=redefined-builtin
if isinstance(new, float):
format = lambda num: "%.3f" % num
old = old_stats.get(m_type)
if old is not None:
diff_str = diff_string(old, new)
old = format(old)
else:
old, diff_str = "NC", "NC"
lines += (m_type.replace("_", " "), format(new), old, diff_str)
return lines
class BaseChecker(OptionsProviderMixIn):
"""base class for checkers"""
# checker name (you may reuse an existing one)
name = None # type: str
# options level (0 will be displaying in --help, 1 in --long-help)
level = 1
# ordered list of options to control the ckecker behaviour
options = () # type: Any
# messages issued by this checker
msgs = {} # type: Any
# reports issued by this checker
reports = () # type: Any
# mark this checker as enabled or not.
enabled = True
def __init__(self, linter=None):
"""checker instances should have the linter as argument
linter is an object implementing ILinter
"""
if self.name is not None:
self.name = self.name.lower()
OptionsProviderMixIn.__init__(self)
self.linter = linter
def add_message(
self,
msg_id,
line=None,
node=None,
args=None,
confidence=UNDEFINED,
col_offset=None,
):
"""add a message of a given type"""
self.linter.add_message(msg_id, line, node, args, confidence, col_offset)
# dummy methods implementing the IChecker interface
def open(self):
"""called before visiting project (i.e set of modules)"""
def close(self):
"""called after visiting project (i.e set of modules)"""
class BaseTokenChecker(BaseChecker):
"""Base class for checkers that want to have access to the token stream."""
def process_tokens(self, tokens):
"""Should be overridden by subclasses."""
raise NotImplementedError()
def initialize(linter):
"""initialize linter with checkers in this package """
register_plugins(linter, __path__[0])
__all__ = ("BaseChecker", "BaseTokenChecker", "initialize")

View File

@@ -0,0 +1,92 @@
# Copyright (c) 2015-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2017 Derek Gustafson <degustaf@gmail.com>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""Checker for anything related to the async protocol (PEP 492)."""
import sys
import astroid
from astroid import bases
from astroid import exceptions
from pylint import checkers
from pylint.checkers import utils as checker_utils
from pylint import interfaces
from pylint import utils
from pylint.checkers.utils import decorated_with
class AsyncChecker(checkers.BaseChecker):
__implements__ = interfaces.IAstroidChecker
name = "async"
msgs = {
"E1700": (
"Yield inside async function",
"yield-inside-async-function",
"Used when an `yield` or `yield from` statement is "
"found inside an async function.",
{"minversion": (3, 5)},
),
"E1701": (
"Async context manager '%s' doesn't implement __aenter__ and __aexit__.",
"not-async-context-manager",
"Used when an async context manager is used with an object "
"that does not implement the async context management protocol.",
{"minversion": (3, 5)},
),
}
def open(self):
self._ignore_mixin_members = utils.get_global_option(
self, "ignore-mixin-members"
)
self._async_generators = ["contextlib.asynccontextmanager"]
@checker_utils.check_messages("yield-inside-async-function")
def visit_asyncfunctiondef(self, node):
for child in node.nodes_of_class(astroid.Yield):
if child.scope() is node and (
sys.version_info[:2] == (3, 5) or isinstance(child, astroid.YieldFrom)
):
self.add_message("yield-inside-async-function", node=child)
@checker_utils.check_messages("not-async-context-manager")
def visit_asyncwith(self, node):
for ctx_mgr, _ in node.items:
inferred = checker_utils.safe_infer(ctx_mgr)
if inferred is None or inferred is astroid.Uninferable:
continue
if isinstance(inferred, bases.AsyncGenerator):
# Check if we are dealing with a function decorated
# with contextlib.asynccontextmanager.
if decorated_with(inferred.parent, self._async_generators):
continue
else:
try:
inferred.getattr("__aenter__")
inferred.getattr("__aexit__")
except exceptions.NotFoundError:
if isinstance(inferred, astroid.Instance):
# If we do not know the bases of this class,
# just skip it.
if not checker_utils.has_known_bases(inferred):
continue
# Just ignore mixin classes.
if self._ignore_mixin_members:
if inferred.name[-5:].lower() == "mixin":
continue
else:
continue
self.add_message(
"not-async-context-manager", node=node, args=(inferred.name,)
)
def register(linter):
"""required method to auto register this checker"""
linter.register_checker(AsyncChecker(linter))

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,530 @@
# -*- coding: utf-8 -*-
# Copyright (c) 2006, 2009-2010, 2012-2015 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2012, 2014 Google, Inc.
# Copyright (c) 2014-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Łukasz Rogalski <rogalski.91@gmail.com>
# Copyright (c) 2017 ahirnish <ahirnish@gmail.com>
# Copyright (c) 2018 Mike Frysinger <vapier@gmail.com>
# Copyright (c) 2018 Mark Miller <725mrm@gmail.com>
# Copyright (c) 2018 Ashley Whetter <ashley@awhetter.co.uk>
# Copyright (c) 2018 Ville Skyttä <ville.skytta@upcloud.com>
# Copyright (c) 2018 Jakub Wilk <jwilk@jwilk.net>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""check for signs of poor design"""
from collections import defaultdict
import re
import astroid
from astroid import If, BoolOp
from astroid import decorators
from pylint.interfaces import IAstroidChecker
from pylint.checkers import BaseChecker
from pylint.checkers.utils import check_messages
from pylint import utils
MSGS = {
"R0901": (
"Too many ancestors (%s/%s)",
"too-many-ancestors",
"Used when class has too many parent classes, try to reduce "
"this to get a simpler (and so easier to use) class.",
),
"R0902": (
"Too many instance attributes (%s/%s)",
"too-many-instance-attributes",
"Used when class has too many instance attributes, try to reduce "
"this to get a simpler (and so easier to use) class.",
),
"R0903": (
"Too few public methods (%s/%s)",
"too-few-public-methods",
"Used when class has too few public methods, so be sure it's "
"really worth it.",
),
"R0904": (
"Too many public methods (%s/%s)",
"too-many-public-methods",
"Used when class has too many public methods, try to reduce "
"this to get a simpler (and so easier to use) class.",
),
"R0911": (
"Too many return statements (%s/%s)",
"too-many-return-statements",
"Used when a function or method has too many return statement, "
"making it hard to follow.",
),
"R0912": (
"Too many branches (%s/%s)",
"too-many-branches",
"Used when a function or method has too many branches, "
"making it hard to follow.",
),
"R0913": (
"Too many arguments (%s/%s)",
"too-many-arguments",
"Used when a function or method takes too many arguments.",
),
"R0914": (
"Too many local variables (%s/%s)",
"too-many-locals",
"Used when a function or method has too many local variables.",
),
"R0915": (
"Too many statements (%s/%s)",
"too-many-statements",
"Used when a function or method has too many statements. You "
"should then split it in smaller functions / methods.",
),
"R0916": (
"Too many boolean expressions in if statement (%s/%s)",
"too-many-boolean-expressions",
"Used when an if statement contains too many boolean expressions.",
),
}
SPECIAL_OBJ = re.compile("^_{2}[a-z]+_{2}$")
DATACLASS_DECORATOR = "dataclass"
DATACLASS_IMPORT = "dataclasses"
TYPING_NAMEDTUPLE = "typing.NamedTuple"
def _is_typing_namedtuple(node: astroid.ClassDef) -> bool:
"""Check if a class node is a typing.NamedTuple class"""
for base in node.ancestors():
if base.qname() == TYPING_NAMEDTUPLE:
return True
return False
def _is_enum_class(node: astroid.ClassDef) -> bool:
"""Check if a class definition defines an Enum class.
:param node: The class node to check.
:type node: astroid.ClassDef
:returns: True if the given node represents an Enum class. False otherwise.
:rtype: bool
"""
for base in node.bases:
try:
inferred_bases = base.inferred()
except astroid.InferenceError:
continue
for ancestor in inferred_bases:
if not isinstance(ancestor, astroid.ClassDef):
continue
if ancestor.name == "Enum" and ancestor.root().name == "enum":
return True
return False
def _is_dataclass(node: astroid.ClassDef) -> bool:
"""Check if a class definition defines a Python 3.7+ dataclass
:param node: The class node to check.
:type node: astroid.ClassDef
:returns: True if the given node represents a dataclass class. False otherwise.
:rtype: bool
"""
if not node.decorators:
return False
root_locals = node.root().locals
for decorator in node.decorators.nodes:
if isinstance(decorator, astroid.Call):
decorator = decorator.func
if not isinstance(decorator, (astroid.Name, astroid.Attribute)):
continue
if isinstance(decorator, astroid.Name):
name = decorator.name
else:
name = decorator.attrname
if name == DATACLASS_DECORATOR and DATACLASS_DECORATOR in root_locals:
return True
return False
def _count_boolean_expressions(bool_op):
"""Counts the number of boolean expressions in BoolOp `bool_op` (recursive)
example: a and (b or c or (d and e)) ==> 5 boolean expressions
"""
nb_bool_expr = 0
for bool_expr in bool_op.get_children():
if isinstance(bool_expr, BoolOp):
nb_bool_expr += _count_boolean_expressions(bool_expr)
else:
nb_bool_expr += 1
return nb_bool_expr
def _count_methods_in_class(node):
all_methods = sum(1 for method in node.methods() if not method.name.startswith("_"))
# Special methods count towards the number of public methods,
# but don't count towards there being too many methods.
for method in node.mymethods():
if SPECIAL_OBJ.search(method.name) and method.name != "__init__":
all_methods += 1
return all_methods
class MisdesignChecker(BaseChecker):
"""checks for sign of poor/misdesign:
* number of methods, attributes, local variables...
* size, complexity of functions, methods
"""
__implements__ = (IAstroidChecker,)
# configuration section name
name = "design"
# messages
msgs = MSGS
priority = -2
# configuration options
options = (
(
"max-args",
{
"default": 5,
"type": "int",
"metavar": "<int>",
"help": "Maximum number of arguments for function / method.",
},
),
(
"max-locals",
{
"default": 15,
"type": "int",
"metavar": "<int>",
"help": "Maximum number of locals for function / method body.",
},
),
(
"max-returns",
{
"default": 6,
"type": "int",
"metavar": "<int>",
"help": "Maximum number of return / yield for function / "
"method body.",
},
),
(
"max-branches",
{
"default": 12,
"type": "int",
"metavar": "<int>",
"help": "Maximum number of branch for function / method body.",
},
),
(
"max-statements",
{
"default": 50,
"type": "int",
"metavar": "<int>",
"help": "Maximum number of statements in function / method " "body.",
},
),
(
"max-parents",
{
"default": 7,
"type": "int",
"metavar": "<num>",
"help": "Maximum number of parents for a class (see R0901).",
},
),
(
"max-attributes",
{
"default": 7,
"type": "int",
"metavar": "<num>",
"help": "Maximum number of attributes for a class \
(see R0902).",
},
),
(
"min-public-methods",
{
"default": 2,
"type": "int",
"metavar": "<num>",
"help": "Minimum number of public methods for a class \
(see R0903).",
},
),
(
"max-public-methods",
{
"default": 20,
"type": "int",
"metavar": "<num>",
"help": "Maximum number of public methods for a class \
(see R0904).",
},
),
(
"max-bool-expr",
{
"default": 5,
"type": "int",
"metavar": "<num>",
"help": "Maximum number of boolean expressions in an if " "statement.",
},
),
)
def __init__(self, linter=None):
BaseChecker.__init__(self, linter)
self.stats = None
self._returns = None
self._branches = None
self._stmts = None
def open(self):
"""initialize visit variables"""
self.stats = self.linter.add_stats()
self._returns = []
self._branches = defaultdict(int)
self._stmts = []
def _inc_all_stmts(self, amount):
for i in range(len(self._stmts)):
self._stmts[i] += amount
@decorators.cachedproperty
def _ignored_argument_names(self):
return utils.get_global_option(self, "ignored-argument-names", default=None)
@check_messages(
"too-many-ancestors",
"too-many-instance-attributes",
"too-few-public-methods",
"too-many-public-methods",
)
def visit_classdef(self, node):
"""check size of inheritance hierarchy and number of instance attributes
"""
nb_parents = len(list(node.ancestors()))
if nb_parents > self.config.max_parents:
self.add_message(
"too-many-ancestors",
node=node,
args=(nb_parents, self.config.max_parents),
)
if len(node.instance_attrs) > self.config.max_attributes:
self.add_message(
"too-many-instance-attributes",
node=node,
args=(len(node.instance_attrs), self.config.max_attributes),
)
@check_messages("too-few-public-methods", "too-many-public-methods")
def leave_classdef(self, node):
"""check number of public methods"""
my_methods = sum(
1 for method in node.mymethods() if not method.name.startswith("_")
)
# Does the class contain less than n public methods ?
# This checks only the methods defined in the current class,
# since the user might not have control over the classes
# from the ancestors. It avoids some false positives
# for classes such as unittest.TestCase, which provides
# a lot of assert methods. It doesn't make sense to warn
# when the user subclasses TestCase to add his own tests.
if my_methods > self.config.max_public_methods:
self.add_message(
"too-many-public-methods",
node=node,
args=(my_methods, self.config.max_public_methods),
)
# Stop here for exception, metaclass, interface classes and other
# classes for which we don't need to count the methods.
if (
node.type != "class"
or _is_enum_class(node)
or _is_dataclass(node)
or _is_typing_namedtuple(node)
):
return
# Does the class contain more than n public methods ?
# This checks all the methods defined by ancestors and
# by the current class.
all_methods = _count_methods_in_class(node)
if all_methods < self.config.min_public_methods:
self.add_message(
"too-few-public-methods",
node=node,
args=(all_methods, self.config.min_public_methods),
)
@check_messages(
"too-many-return-statements",
"too-many-branches",
"too-many-arguments",
"too-many-locals",
"too-many-statements",
"keyword-arg-before-vararg",
)
def visit_functiondef(self, node):
"""check function name, docstring, arguments, redefinition,
variable names, max locals
"""
# init branch and returns counters
self._returns.append(0)
# check number of arguments
args = node.args.args
ignored_argument_names = self._ignored_argument_names
if args is not None:
ignored_args_num = 0
if ignored_argument_names:
ignored_args_num = sum(
1 for arg in args if ignored_argument_names.match(arg.name)
)
argnum = len(args) - ignored_args_num
if argnum > self.config.max_args:
self.add_message(
"too-many-arguments",
node=node,
args=(len(args), self.config.max_args),
)
else:
ignored_args_num = 0
# check number of local variables
locnum = len(node.locals) - ignored_args_num
if locnum > self.config.max_locals:
self.add_message(
"too-many-locals", node=node, args=(locnum, self.config.max_locals)
)
# init new statements counter
self._stmts.append(1)
visit_asyncfunctiondef = visit_functiondef
@check_messages(
"too-many-return-statements",
"too-many-branches",
"too-many-arguments",
"too-many-locals",
"too-many-statements",
)
def leave_functiondef(self, node):
"""most of the work is done here on close:
checks for max returns, branch, return in __init__
"""
returns = self._returns.pop()
if returns > self.config.max_returns:
self.add_message(
"too-many-return-statements",
node=node,
args=(returns, self.config.max_returns),
)
branches = self._branches[node]
if branches > self.config.max_branches:
self.add_message(
"too-many-branches",
node=node,
args=(branches, self.config.max_branches),
)
# check number of statements
stmts = self._stmts.pop()
if stmts > self.config.max_statements:
self.add_message(
"too-many-statements",
node=node,
args=(stmts, self.config.max_statements),
)
leave_asyncfunctiondef = leave_functiondef
def visit_return(self, _):
"""count number of returns"""
if not self._returns:
return # return outside function, reported by the base checker
self._returns[-1] += 1
def visit_default(self, node):
"""default visit method -> increments the statements counter if
necessary
"""
if node.is_statement:
self._inc_all_stmts(1)
def visit_tryexcept(self, node):
"""increments the branches counter"""
branches = len(node.handlers)
if node.orelse:
branches += 1
self._inc_branch(node, branches)
self._inc_all_stmts(branches)
def visit_tryfinally(self, node):
"""increments the branches counter"""
self._inc_branch(node, 2)
self._inc_all_stmts(2)
@check_messages("too-many-boolean-expressions")
def visit_if(self, node):
"""increments the branches counter and checks boolean expressions"""
self._check_boolean_expressions(node)
branches = 1
# don't double count If nodes coming from some 'elif'
if node.orelse and (len(node.orelse) > 1 or not isinstance(node.orelse[0], If)):
branches += 1
self._inc_branch(node, branches)
self._inc_all_stmts(branches)
def _check_boolean_expressions(self, node):
"""Go through "if" node `node` and counts its boolean expressions
if the "if" node test is a BoolOp node
"""
condition = node.test
if not isinstance(condition, BoolOp):
return
nb_bool_expr = _count_boolean_expressions(condition)
if nb_bool_expr > self.config.max_bool_expr:
self.add_message(
"too-many-boolean-expressions",
node=condition,
args=(nb_bool_expr, self.config.max_bool_expr),
)
def visit_while(self, node):
"""increments the branches counter"""
branches = 1
if node.orelse:
branches += 1
self._inc_branch(node, branches)
visit_for = visit_while
def _inc_branch(self, node, branchesnum=1):
"""increments the branches counter"""
self._branches[node.scope()] += branchesnum
def register(linter):
"""required method to auto register this checker """
linter.register_checker(MisdesignChecker(linter))

View File

@@ -0,0 +1,561 @@
# -*- coding: utf-8 -*-
# Copyright (c) 2006-2011, 2013-2014 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2011-2014 Google, Inc.
# Copyright (c) 2012 Tim Hatch <tim@timhatch.com>
# Copyright (c) 2013-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Brett Cannon <brett@python.org>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015 Rene Zhang <rz99@cornell.edu>
# Copyright (c) 2015 Florian Bruhin <me@the-compiler.org>
# Copyright (c) 2015 Steven Myint <hg@stevenmyint.com>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Erik <erik.eriksson@yahoo.com>
# Copyright (c) 2016 Jakub Wilk <jwilk@jwilk.net>
# Copyright (c) 2017 Łukasz Rogalski <rogalski.91@gmail.com>
# Copyright (c) 2017 Martin von Gagern <gagern@google.com>
# Copyright (c) 2018 Mike Frysinger <vapier@gmail.com>
# Copyright (c) 2018 ssolanki <sushobhitsolanki@gmail.com>
# Copyright (c) 2018 Alexander Todorov <atodorov@otb.bg>
# Copyright (c) 2018 Ville Skyttä <ville.skytta@upcloud.com>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""Checks for various exception related errors."""
import builtins
import inspect
import sys
import typing
import astroid
from pylint import checkers
from pylint.checkers import utils
from pylint import interfaces
def _builtin_exceptions():
def predicate(obj):
return isinstance(obj, type) and issubclass(obj, BaseException)
members = inspect.getmembers(builtins, predicate)
return {exc.__name__ for (_, exc) in members}
def _annotated_unpack_infer(stmt, context=None):
"""
Recursively generate nodes inferred by the given statement.
If the inferred value is a list or a tuple, recurse on the elements.
Returns an iterator which yields tuples in the format
('original node', 'infered node').
"""
if isinstance(stmt, (astroid.List, astroid.Tuple)):
for elt in stmt.elts:
inferred = utils.safe_infer(elt)
if inferred and inferred is not astroid.Uninferable:
yield elt, inferred
return
for infered in stmt.infer(context):
if infered is astroid.Uninferable:
continue
yield stmt, infered
def _is_raising(body: typing.List) -> bool:
"""Return true if the given statement node raise an exception"""
for node in body:
if isinstance(node, astroid.Raise):
return True
return False
PY3K = sys.version_info >= (3, 0)
OVERGENERAL_EXCEPTIONS = ("BaseException", "Exception")
BUILTINS_NAME = builtins.__name__
MSGS = {
"E0701": (
"Bad except clauses order (%s)",
"bad-except-order",
"Used when except clauses are not in the correct order (from the "
"more specific to the more generic). If you don't fix the order, "
"some exceptions may not be caught by the most specific handler.",
),
"E0702": (
"Raising %s while only classes or instances are allowed",
"raising-bad-type",
"Used when something which is neither a class, an instance or a "
"string is raised (i.e. a `TypeError` will be raised).",
),
"E0703": (
"Exception context set to something which is not an exception, nor None",
"bad-exception-context",
'Used when using the syntax "raise ... from ...", '
"where the exception context is not an exception, "
"nor None.",
),
"E0704": (
"The raise statement is not inside an except clause",
"misplaced-bare-raise",
"Used when a bare raise is not used inside an except clause. "
"This generates an error, since there are no active exceptions "
"to be reraised. An exception to this rule is represented by "
"a bare raise inside a finally clause, which might work, as long "
"as an exception is raised inside the try block, but it is "
"nevertheless a code smell that must not be relied upon.",
),
"E0710": (
"Raising a new style class which doesn't inherit from BaseException",
"raising-non-exception",
"Used when a new style class which doesn't inherit from "
"BaseException is raised.",
),
"E0711": (
"NotImplemented raised - should raise NotImplementedError",
"notimplemented-raised",
"Used when NotImplemented is raised instead of NotImplementedError",
),
"E0712": (
"Catching an exception which doesn't inherit from Exception: %s",
"catching-non-exception",
"Used when a class which doesn't inherit from "
"Exception is used as an exception in an except clause.",
),
"W0702": (
"No exception type(s) specified",
"bare-except",
"Used when an except clause doesn't specify exceptions type to catch.",
),
"W0703": (
"Catching too general exception %s",
"broad-except",
"Used when an except catches a too general exception, "
"possibly burying unrelated errors.",
),
"W0705": (
"Catching previously caught exception type %s",
"duplicate-except",
"Used when an except catches a type that was already caught by "
"a previous handler.",
),
"W0706": (
"The except handler raises immediately",
"try-except-raise",
"Used when an except handler uses raise as its first or only "
"operator. This is useless because it raises back the exception "
"immediately. Remove the raise operator or the entire "
"try-except-raise block!",
),
"W0711": (
'Exception to catch is the result of a binary "%s" operation',
"binary-op-exception",
"Used when the exception to catch is of the form "
'"except A or B:". If intending to catch multiple, '
'rewrite as "except (A, B):"',
),
"W0715": (
"Exception arguments suggest string formatting might be intended",
"raising-format-tuple",
"Used when passing multiple arguments to an exception "
"constructor, the first of them a string literal containing what "
"appears to be placeholders intended for formatting",
),
"W0716": (
"Invalid exception operation. %s",
"wrong-exception-operation",
"Used when an operation is done against an exception, but the operation "
"is not valid for the exception in question. Usually emitted when having "
"binary operations between exceptions in except handlers.",
),
}
class BaseVisitor:
"""Base class for visitors defined in this module."""
def __init__(self, checker, node):
self._checker = checker
self._node = node
def visit(self, node):
name = node.__class__.__name__.lower()
dispatch_meth = getattr(self, "visit_" + name, None)
if dispatch_meth:
dispatch_meth(node)
else:
self.visit_default(node)
def visit_default(self, node): # pylint: disable=unused-argument
"""Default implementation for all the nodes."""
class ExceptionRaiseRefVisitor(BaseVisitor):
"""Visit references (anything that is not an AST leaf)."""
def visit_name(self, name):
if name.name == "NotImplemented":
self._checker.add_message("notimplemented-raised", node=self._node)
def visit_call(self, call):
if isinstance(call.func, astroid.Name):
self.visit_name(call.func)
if (
len(call.args) > 1
and isinstance(call.args[0], astroid.Const)
and isinstance(call.args[0].value, str)
):
msg = call.args[0].value
if "%" in msg or ("{" in msg and "}" in msg):
self._checker.add_message("raising-format-tuple", node=self._node)
class ExceptionRaiseLeafVisitor(BaseVisitor):
"""Visitor for handling leaf kinds of a raise value."""
def visit_const(self, const):
if not isinstance(const.value, str):
# raising-string will be emitted from python3 porting checker.
self._checker.add_message(
"raising-bad-type", node=self._node, args=const.value.__class__.__name__
)
def visit_instance(self, instance):
# pylint: disable=protected-access
cls = instance._proxied
self.visit_classdef(cls)
# Exception instances have a particular class type
visit_exceptioninstance = visit_instance
def visit_classdef(self, cls):
if not utils.inherit_from_std_ex(cls) and utils.has_known_bases(cls):
if cls.newstyle:
self._checker.add_message("raising-non-exception", node=self._node)
else:
self._checker.add_message("nonstandard-exception", node=self._node)
def visit_tuple(self, tuple_node):
if PY3K or not tuple_node.elts:
self._checker.add_message("raising-bad-type", node=self._node, args="tuple")
return
# On Python 2, using the following is not an error:
# raise (ZeroDivisionError, None)
# raise (ZeroDivisionError, )
# What's left to do is to check that the first
# argument is indeed an exception. Verifying the other arguments
# is not the scope of this check.
first = tuple_node.elts[0]
inferred = utils.safe_infer(first)
if not inferred or inferred is astroid.Uninferable:
return
if (
isinstance(inferred, astroid.Instance)
and inferred.__class__.__name__ != "Instance"
):
# TODO: explain why
self.visit_default(tuple_node)
else:
self.visit(inferred)
def visit_default(self, node):
name = getattr(node, "name", node.__class__.__name__)
self._checker.add_message("raising-bad-type", node=self._node, args=name)
class ExceptionsChecker(checkers.BaseChecker):
"""Exception related checks."""
__implements__ = interfaces.IAstroidChecker
name = "exceptions"
msgs = MSGS
priority = -4
options = (
(
"overgeneral-exceptions",
{
"default": OVERGENERAL_EXCEPTIONS,
"type": "csv",
"metavar": "<comma-separated class names>",
"help": "Exceptions that will emit a warning "
'when being caught. Defaults to "%s".'
% (", ".join(OVERGENERAL_EXCEPTIONS),),
},
),
)
def open(self):
self._builtin_exceptions = _builtin_exceptions()
super(ExceptionsChecker, self).open()
@utils.check_messages(
"nonstandard-exception",
"misplaced-bare-raise",
"raising-bad-type",
"raising-non-exception",
"notimplemented-raised",
"bad-exception-context",
"raising-format-tuple",
)
def visit_raise(self, node):
if node.exc is None:
self._check_misplaced_bare_raise(node)
return
if PY3K and node.cause:
self._check_bad_exception_context(node)
expr = node.exc
ExceptionRaiseRefVisitor(self, node).visit(expr)
try:
inferred_value = expr.inferred()[-1]
except astroid.InferenceError:
pass
else:
if inferred_value:
ExceptionRaiseLeafVisitor(self, node).visit(inferred_value)
def _check_misplaced_bare_raise(self, node):
# Filter out if it's present in __exit__.
scope = node.scope()
if (
isinstance(scope, astroid.FunctionDef)
and scope.is_method()
and scope.name == "__exit__"
):
return
current = node
# Stop when a new scope is generated or when the raise
# statement is found inside a TryFinally.
ignores = (astroid.ExceptHandler, astroid.FunctionDef)
while current and not isinstance(current.parent, ignores):
current = current.parent
expected = (astroid.ExceptHandler,)
if not current or not isinstance(current.parent, expected):
self.add_message("misplaced-bare-raise", node=node)
def _check_bad_exception_context(self, node):
"""Verify that the exception context is properly set.
An exception context can be only `None` or an exception.
"""
cause = utils.safe_infer(node.cause)
if cause in (astroid.Uninferable, None):
return
if isinstance(cause, astroid.Const):
if cause.value is not None:
self.add_message("bad-exception-context", node=node)
elif not isinstance(cause, astroid.ClassDef) and not utils.inherit_from_std_ex(
cause
):
self.add_message("bad-exception-context", node=node)
def _check_catching_non_exception(self, handler, exc, part):
if isinstance(exc, astroid.Tuple):
# Check if it is a tuple of exceptions.
inferred = [utils.safe_infer(elt) for elt in exc.elts]
if any(node is astroid.Uninferable for node in inferred):
# Don't emit if we don't know every component.
return
if all(
node
and (utils.inherit_from_std_ex(node) or not utils.has_known_bases(node))
for node in inferred
):
return
if not isinstance(exc, astroid.ClassDef):
# Don't emit the warning if the infered stmt
# is None, but the exception handler is something else,
# maybe it was redefined.
if isinstance(exc, astroid.Const) and exc.value is None:
if (
isinstance(handler.type, astroid.Const)
and handler.type.value is None
) or handler.type.parent_of(exc):
# If the exception handler catches None or
# the exception component, which is None, is
# defined by the entire exception handler, then
# emit a warning.
self.add_message(
"catching-non-exception",
node=handler.type,
args=(part.as_string(),),
)
else:
self.add_message(
"catching-non-exception",
node=handler.type,
args=(part.as_string(),),
)
return
if (
not utils.inherit_from_std_ex(exc)
and exc.name not in self._builtin_exceptions
):
if utils.has_known_bases(exc):
self.add_message(
"catching-non-exception", node=handler.type, args=(exc.name,)
)
def _check_try_except_raise(self, node):
def gather_exceptions_from_handler(handler):
exceptions = []
if handler.type:
exceptions_in_handler = utils.safe_infer(handler.type)
if isinstance(exceptions_in_handler, astroid.Tuple):
exceptions = {
exception
for exception in exceptions_in_handler.elts
if isinstance(exception, astroid.Name)
}
elif exceptions_in_handler:
exceptions = [exceptions_in_handler]
return exceptions
bare_raise = False
handler_having_bare_raise = None
excs_in_bare_handler = []
for handler in node.handlers:
if bare_raise:
# check that subsequent handler is not parent of handler which had bare raise.
# since utils.safe_infer can fail for bare except, check it before.
# also break early if bare except is followed by bare except.
excs_in_current_handler = gather_exceptions_from_handler(handler)
if not excs_in_current_handler:
bare_raise = False
break
for exc_in_current_handler in excs_in_current_handler:
inferred_current = utils.safe_infer(exc_in_current_handler)
if any(
utils.is_subclass_of(
utils.safe_infer(exc_in_bare_handler), inferred_current
)
for exc_in_bare_handler in excs_in_bare_handler
):
bare_raise = False
break
# `raise` as the first operator inside the except handler
if _is_raising([handler.body[0]]):
# flags when there is a bare raise
if handler.body[0].exc is None:
bare_raise = True
handler_having_bare_raise = handler
excs_in_bare_handler = gather_exceptions_from_handler(handler)
if bare_raise:
self.add_message("try-except-raise", node=handler_having_bare_raise)
@utils.check_messages("wrong-exception-operation")
def visit_binop(self, node):
if isinstance(node.parent, astroid.ExceptHandler):
# except (V | A)
suggestion = "Did you mean '(%s, %s)' instead?" % (
node.left.as_string(),
node.right.as_string(),
)
self.add_message("wrong-exception-operation", node=node, args=(suggestion,))
@utils.check_messages("wrong-exception-operation")
def visit_compare(self, node):
if isinstance(node.parent, astroid.ExceptHandler):
# except (V < A)
suggestion = "Did you mean '(%s, %s)' instead?" % (
node.left.as_string(),
", ".join(operand.as_string() for _, operand in node.ops),
)
self.add_message("wrong-exception-operation", node=node, args=(suggestion,))
@utils.check_messages(
"bare-except",
"broad-except",
"try-except-raise",
"binary-op-exception",
"bad-except-order",
"catching-non-exception",
"duplicate-except",
)
def visit_tryexcept(self, node):
"""check for empty except"""
self._check_try_except_raise(node)
exceptions_classes = []
nb_handlers = len(node.handlers)
for index, handler in enumerate(node.handlers):
if handler.type is None:
if not _is_raising(handler.body):
self.add_message("bare-except", node=handler)
# check if an "except:" is followed by some other
# except
if index < (nb_handlers - 1):
msg = "empty except clause should always appear last"
self.add_message("bad-except-order", node=node, args=msg)
elif isinstance(handler.type, astroid.BoolOp):
self.add_message(
"binary-op-exception", node=handler, args=handler.type.op
)
else:
try:
excs = list(_annotated_unpack_infer(handler.type))
except astroid.InferenceError:
continue
for part, exc in excs:
if exc is astroid.Uninferable:
continue
if isinstance(exc, astroid.Instance) and utils.inherit_from_std_ex(
exc
):
# pylint: disable=protected-access
exc = exc._proxied
self._check_catching_non_exception(handler, exc, part)
if not isinstance(exc, astroid.ClassDef):
continue
exc_ancestors = [
anc
for anc in exc.ancestors()
if isinstance(anc, astroid.ClassDef)
]
for previous_exc in exceptions_classes:
if previous_exc in exc_ancestors:
msg = "%s is an ancestor class of %s" % (
previous_exc.name,
exc.name,
)
self.add_message(
"bad-except-order", node=handler.type, args=msg
)
if (
exc.name in self.config.overgeneral_exceptions
and exc.root().name == utils.EXCEPTIONS_MODULE
and not _is_raising(handler.body)
):
self.add_message(
"broad-except", args=exc.name, node=handler.type
)
if exc in exceptions_classes:
self.add_message(
"duplicate-except", args=exc.name, node=handler.type
)
exceptions_classes += [exc for _, exc in excs]
def register(linter):
"""required method to auto register this checker"""
linter.register_checker(ExceptionsChecker(linter))

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,934 @@
# -*- coding: utf-8 -*-
# Copyright (c) 2006-2015 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2012-2014 Google, Inc.
# Copyright (c) 2013 buck@yelp.com <buck@yelp.com>
# Copyright (c) 2014-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Brett Cannon <brett@python.org>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015-2016 Moises Lopez <moylop260@vauxoo.com>
# Copyright (c) 2015 Dmitry Pribysh <dmand@yandex.ru>
# Copyright (c) 2015 Cezar <celnazli@bitdefender.com>
# Copyright (c) 2015 Florian Bruhin <me@the-compiler.org>
# Copyright (c) 2015 Noam Yorav-Raphael <noamraph@gmail.com>
# Copyright (c) 2015 James Morgensen <james.morgensen@gmail.com>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Jared Garst <cultofjared@gmail.com>
# Copyright (c) 2016 Maik Röder <maikroeder@gmail.com>
# Copyright (c) 2016 Glenn Matthews <glenn@e-dad.net>
# Copyright (c) 2016 Ashley Whetter <ashley@awhetter.co.uk>
# Copyright (c) 2017 hippo91 <guillaume.peillex@gmail.com>
# Copyright (c) 2017 Michka Popoff <michkapopoff@gmail.com>
# Copyright (c) 2017 Łukasz Rogalski <rogalski.91@gmail.com>
# Copyright (c) 2017 Erik Wright <erik.wright@shopify.com>
# Copyright (c) 2018 Mike Frysinger <vapier@gmail.com>
# Copyright (c) 2018 Sushobhit <31987769+sushobhit27@users.noreply.github.com>
# Copyright (c) 2018 Marianna Polatoglou <mpolatoglou@bloomberg.net>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""imports checkers for Python code"""
import collections
from distutils import sysconfig
import os
import sys
import copy
import astroid
from astroid import are_exclusive, decorators
from astroid.modutils import get_module_part, is_standard_module
import isort
from pylint.interfaces import IAstroidChecker
from pylint.utils import get_global_option
from pylint.exceptions import EmptyReportError
from pylint.checkers import BaseChecker
from pylint.checkers.utils import (
check_messages,
node_ignores_exception,
is_from_fallback_block,
)
from pylint.graph import get_cycles, DotBackend
from pylint.reporters.ureports.nodes import VerbatimText, Paragraph
def _qualified_names(modname):
"""Split the names of the given module into subparts
For example,
_qualified_names('pylint.checkers.ImportsChecker')
returns
['pylint', 'pylint.checkers', 'pylint.checkers.ImportsChecker']
"""
names = modname.split(".")
return [".".join(names[0 : i + 1]) for i in range(len(names))]
def _get_import_name(importnode, modname):
"""Get a prepared module name from the given import node
In the case of relative imports, this will return the
absolute qualified module name, which might be useful
for debugging. Otherwise, the initial module name
is returned unchanged.
"""
if isinstance(importnode, astroid.ImportFrom):
if importnode.level:
root = importnode.root()
if isinstance(root, astroid.Module):
modname = root.relative_to_absolute_name(
modname, level=importnode.level
)
return modname
def _get_first_import(node, context, name, base, level, alias):
"""return the node where [base.]<name> is imported or None if not found
"""
fullname = "%s.%s" % (base, name) if base else name
first = None
found = False
for first in context.body:
if first is node:
continue
if first.scope() is node.scope() and first.fromlineno > node.fromlineno:
continue
if isinstance(first, astroid.Import):
if any(fullname == iname[0] for iname in first.names):
found = True
break
elif isinstance(first, astroid.ImportFrom):
if level == first.level:
for imported_name, imported_alias in first.names:
if fullname == "%s.%s" % (first.modname, imported_name):
found = True
break
if (
name != "*"
and name == imported_name
and not (alias or imported_alias)
):
found = True
break
if found:
break
if found and not are_exclusive(first, node):
return first
return None
def _ignore_import_failure(node, modname, ignored_modules):
for submodule in _qualified_names(modname):
if submodule in ignored_modules:
return True
return node_ignores_exception(node, ImportError)
# utilities to represents import dependencies as tree and dot graph ###########
def _make_tree_defs(mod_files_list):
"""get a list of 2-uple (module, list_of_files_which_import_this_module),
it will return a dictionary to represent this as a tree
"""
tree_defs = {}
for mod, files in mod_files_list:
node = (tree_defs, ())
for prefix in mod.split("."):
node = node[0].setdefault(prefix, [{}, []])
node[1] += files
return tree_defs
def _repr_tree_defs(data, indent_str=None):
"""return a string which represents imports as a tree"""
lines = []
nodes = data.items()
for i, (mod, (sub, files)) in enumerate(sorted(nodes, key=lambda x: x[0])):
if not files:
files = ""
else:
files = "(%s)" % ",".join(sorted(files))
if indent_str is None:
lines.append("%s %s" % (mod, files))
sub_indent_str = " "
else:
lines.append(r"%s\-%s %s" % (indent_str, mod, files))
if i == len(nodes) - 1:
sub_indent_str = "%s " % indent_str
else:
sub_indent_str = "%s| " % indent_str
if sub:
lines.append(_repr_tree_defs(sub, sub_indent_str))
return "\n".join(lines)
def _dependencies_graph(filename, dep_info):
"""write dependencies as a dot (graphviz) file
"""
done = {}
printer = DotBackend(filename[:-4], rankdir="LR")
printer.emit('URL="." node[shape="box"]')
for modname, dependencies in sorted(dep_info.items()):
done[modname] = 1
printer.emit_node(modname)
for depmodname in dependencies:
if depmodname not in done:
done[depmodname] = 1
printer.emit_node(depmodname)
for depmodname, dependencies in sorted(dep_info.items()):
for modname in dependencies:
printer.emit_edge(modname, depmodname)
printer.generate(filename)
def _make_graph(filename, dep_info, sect, gtype):
"""generate a dependencies graph and add some information about it in the
report's section
"""
_dependencies_graph(filename, dep_info)
sect.append(Paragraph("%simports graph has been written to %s" % (gtype, filename)))
# the import checker itself ###################################################
MSGS = {
"E0401": (
"Unable to import %s",
"import-error",
"Used when pylint has been unable to import a module.",
{"old_names": [("F0401", "import-error")]},
),
"E0402": (
"Attempted relative import beyond top-level package",
"relative-beyond-top-level",
"Used when a relative import tries to access too many levels "
"in the current package.",
),
"R0401": (
"Cyclic import (%s)",
"cyclic-import",
"Used when a cyclic import between two or more modules is detected.",
),
"W0401": (
"Wildcard import %s",
"wildcard-import",
"Used when `from module import *` is detected.",
),
"W0402": (
"Uses of a deprecated module %r",
"deprecated-module",
"Used a module marked as deprecated is imported.",
),
"W0403": (
"Relative import %r, should be %r",
"relative-import",
"Used when an import relative to the package directory is detected.",
{"maxversion": (3, 0)},
),
"W0404": (
"Reimport %r (imported line %s)",
"reimported",
"Used when a module is reimported multiple times.",
),
"W0406": (
"Module import itself",
"import-self",
"Used when a module is importing itself.",
),
"W0410": (
"__future__ import is not the first non docstring statement",
"misplaced-future",
"Python 2.5 and greater require __future__ import to be the "
"first non docstring statement in the module.",
),
"C0410": (
"Multiple imports on one line (%s)",
"multiple-imports",
"Used when import statement importing multiple modules is detected.",
),
"C0411": (
"%s should be placed before %s",
"wrong-import-order",
"Used when PEP8 import order is not respected (standard imports "
"first, then third-party libraries, then local imports)",
),
"C0412": (
"Imports from package %s are not grouped",
"ungrouped-imports",
"Used when imports are not grouped by packages",
),
"C0413": (
'Import "%s" should be placed at the top of the module',
"wrong-import-position",
"Used when code and imports are mixed",
),
"C0414": (
"Import alias does not rename original package",
"useless-import-alias",
"Used when an import alias is same as original package."
"e.g using import numpy as numpy instead of import numpy as np",
),
}
DEFAULT_STANDARD_LIBRARY = ()
DEFAULT_KNOWN_THIRD_PARTY = ("enchant",)
class ImportsChecker(BaseChecker):
"""checks for
* external modules dependencies
* relative / wildcard imports
* cyclic imports
* uses of deprecated modules
"""
__implements__ = IAstroidChecker
name = "imports"
msgs = MSGS
priority = -2
if sys.version_info < (3, 5):
deprecated_modules = ("optparse",)
else:
deprecated_modules = ("optparse", "tkinter.tix")
options = (
(
"deprecated-modules",
{
"default": deprecated_modules,
"type": "csv",
"metavar": "<modules>",
"help": "Deprecated modules which should not be used,"
" separated by a comma.",
},
),
(
"import-graph",
{
"default": "",
"type": "string",
"metavar": "<file.dot>",
"help": "Create a graph of every (i.e. internal and"
" external) dependencies in the given file"
" (report RP0402 must not be disabled).",
},
),
(
"ext-import-graph",
{
"default": "",
"type": "string",
"metavar": "<file.dot>",
"help": "Create a graph of external dependencies in the"
" given file (report RP0402 must not be disabled).",
},
),
(
"int-import-graph",
{
"default": "",
"type": "string",
"metavar": "<file.dot>",
"help": "Create a graph of internal dependencies in the"
" given file (report RP0402 must not be disabled).",
},
),
(
"known-standard-library",
{
"default": DEFAULT_STANDARD_LIBRARY,
"type": "csv",
"metavar": "<modules>",
"help": "Force import order to recognize a module as part of "
"the standard compatibility libraries.",
},
),
(
"known-third-party",
{
"default": DEFAULT_KNOWN_THIRD_PARTY,
"type": "csv",
"metavar": "<modules>",
"help": "Force import order to recognize a module as part of "
"a third party library.",
},
),
(
"analyse-fallback-blocks",
{
"default": False,
"type": "yn",
"metavar": "<y_or_n>",
"help": "Analyse import fallback blocks. This can be used to "
"support both Python 2 and 3 compatible code, which "
"means that the block might have code that exists "
"only in one or another interpreter, leading to false "
"positives when analysed.",
},
),
(
"allow-wildcard-with-all",
{
"default": False,
"type": "yn",
"metavar": "<y_or_n>",
"help": "Allow wildcard imports from modules that define __all__.",
},
),
)
def __init__(self, linter=None):
BaseChecker.__init__(self, linter)
self.stats = None
self.import_graph = None
self._imports_stack = []
self._first_non_import_node = None
self._module_pkg = {} # mapping of modules to the pkg they belong in
self.reports = (
("RP0401", "External dependencies", self._report_external_dependencies),
("RP0402", "Modules dependencies graph", self._report_dependencies_graph),
)
self._site_packages = self._compute_site_packages()
@staticmethod
def _compute_site_packages():
def _normalized_path(path):
return os.path.normcase(os.path.abspath(path))
paths = set()
real_prefix = getattr(sys, "real_prefix", None)
for prefix in filter(None, (real_prefix, sys.prefix)):
path = sysconfig.get_python_lib(prefix=prefix)
path = _normalized_path(path)
paths.add(path)
# Handle Debian's derivatives /usr/local.
if os.path.isfile("/etc/debian_version"):
for prefix in filter(None, (real_prefix, sys.prefix)):
libpython = os.path.join(
prefix,
"local",
"lib",
"python" + sysconfig.get_python_version(),
"dist-packages",
)
paths.add(libpython)
return paths
def open(self):
"""called before visiting project (i.e set of modules)"""
self.linter.add_stats(dependencies={})
self.linter.add_stats(cycles=[])
self.stats = self.linter.stats
self.import_graph = collections.defaultdict(set)
self._module_pkg = {} # mapping of modules to the pkg they belong in
self._excluded_edges = collections.defaultdict(set)
self._ignored_modules = get_global_option(self, "ignored-modules", default=[])
def _import_graph_without_ignored_edges(self):
filtered_graph = copy.deepcopy(self.import_graph)
for node in filtered_graph:
filtered_graph[node].difference_update(self._excluded_edges[node])
return filtered_graph
def close(self):
"""called before visiting project (i.e set of modules)"""
if self.linter.is_message_enabled("cyclic-import"):
graph = self._import_graph_without_ignored_edges()
vertices = list(graph)
for cycle in get_cycles(graph, vertices=vertices):
self.add_message("cyclic-import", args=" -> ".join(cycle))
@check_messages(*MSGS)
def visit_import(self, node):
"""triggered when an import statement is seen"""
self._check_reimport(node)
self._check_import_as_rename(node)
modnode = node.root()
names = [name for name, _ in node.names]
if len(names) >= 2:
self.add_message("multiple-imports", args=", ".join(names), node=node)
for name in names:
self._check_deprecated_module(node, name)
imported_module = self._get_imported_module(node, name)
if isinstance(node.parent, astroid.Module):
# Allow imports nested
self._check_position(node)
if isinstance(node.scope(), astroid.Module):
self._record_import(node, imported_module)
if imported_module is None:
continue
self._check_relative_import(modnode, node, imported_module, name)
self._add_imported_module(node, imported_module.name)
@check_messages(*MSGS)
def visit_importfrom(self, node):
"""triggered when a from statement is seen"""
basename = node.modname
imported_module = self._get_imported_module(node, basename)
self._check_import_as_rename(node)
self._check_misplaced_future(node)
self._check_deprecated_module(node, basename)
self._check_wildcard_imports(node, imported_module)
self._check_same_line_imports(node)
self._check_reimport(node, basename=basename, level=node.level)
if isinstance(node.parent, astroid.Module):
# Allow imports nested
self._check_position(node)
if isinstance(node.scope(), astroid.Module):
self._record_import(node, imported_module)
if imported_module is None:
return
modnode = node.root()
self._check_relative_import(modnode, node, imported_module, basename)
for name, _ in node.names:
if name != "*":
self._add_imported_module(node, "%s.%s" % (imported_module.name, name))
else:
self._add_imported_module(node, imported_module.name)
@check_messages(*MSGS)
def leave_module(self, node):
# Check imports are grouped by category (standard, 3rd party, local)
std_imports, ext_imports, loc_imports = self._check_imports_order(node)
# Check imports are grouped by package within a given category
met = set()
current_package = None
for import_node, import_name in std_imports + ext_imports + loc_imports:
if not self.linter.is_message_enabled(
"ungrouped-imports", import_node.fromlineno
):
continue
package, _, _ = import_name.partition(".")
if current_package and current_package != package and package in met:
self.add_message("ungrouped-imports", node=import_node, args=package)
current_package = package
met.add(package)
self._imports_stack = []
self._first_non_import_node = None
def compute_first_non_import_node(self, node):
if not self.linter.is_message_enabled("wrong-import-position", node.fromlineno):
return
# if the node does not contain an import instruction, and if it is the
# first node of the module, keep a track of it (all the import positions
# of the module will be compared to the position of this first
# instruction)
if self._first_non_import_node:
return
if not isinstance(node.parent, astroid.Module):
return
nested_allowed = [astroid.TryExcept, astroid.TryFinally]
is_nested_allowed = [
allowed for allowed in nested_allowed if isinstance(node, allowed)
]
if is_nested_allowed and any(
node.nodes_of_class((astroid.Import, astroid.ImportFrom))
):
return
if isinstance(node, astroid.Assign):
# Add compatibility for module level dunder names
# https://www.python.org/dev/peps/pep-0008/#module-level-dunder-names
valid_targets = [
isinstance(target, astroid.AssignName)
and target.name.startswith("__")
and target.name.endswith("__")
for target in node.targets
]
if all(valid_targets):
return
self._first_non_import_node = node
visit_tryfinally = (
visit_tryexcept
) = (
visit_assignattr
) = (
visit_assign
) = (
visit_ifexp
) = visit_comprehension = visit_expr = visit_if = compute_first_non_import_node
def visit_functiondef(self, node):
if not self.linter.is_message_enabled("wrong-import-position", node.fromlineno):
return
# If it is the first non import instruction of the module, record it.
if self._first_non_import_node:
return
# Check if the node belongs to an `If` or a `Try` block. If they
# contain imports, skip recording this node.
if not isinstance(node.parent.scope(), astroid.Module):
return
root = node
while not isinstance(root.parent, astroid.Module):
root = root.parent
if isinstance(root, (astroid.If, astroid.TryFinally, astroid.TryExcept)):
if any(root.nodes_of_class((astroid.Import, astroid.ImportFrom))):
return
self._first_non_import_node = node
visit_classdef = visit_for = visit_while = visit_functiondef
def _check_misplaced_future(self, node):
basename = node.modname
if basename == "__future__":
# check if this is the first non-docstring statement in the module
prev = node.previous_sibling()
if prev:
# consecutive future statements are possible
if not (
isinstance(prev, astroid.ImportFrom)
and prev.modname == "__future__"
):
self.add_message("misplaced-future", node=node)
return
def _check_same_line_imports(self, node):
# Detect duplicate imports on the same line.
names = (name for name, _ in node.names)
counter = collections.Counter(names)
for name, count in counter.items():
if count > 1:
self.add_message("reimported", node=node, args=(name, node.fromlineno))
def _check_position(self, node):
"""Check `node` import or importfrom node position is correct
Send a message if `node` comes before another instruction
"""
# if a first non-import instruction has already been encountered,
# it means the import comes after it and therefore is not well placed
if self._first_non_import_node:
self.add_message("wrong-import-position", node=node, args=node.as_string())
def _record_import(self, node, importedmodnode):
"""Record the package `node` imports from"""
if isinstance(node, astroid.ImportFrom):
importedname = node.modname
else:
importedname = importedmodnode.name if importedmodnode else None
if not importedname:
importedname = node.names[0][0].split(".")[0]
if isinstance(node, astroid.ImportFrom) and (node.level or 0) >= 1:
# We need the importedname with first point to detect local package
# Example of node:
# 'from .my_package1 import MyClass1'
# the output should be '.my_package1' instead of 'my_package1'
# Example of node:
# 'from . import my_package2'
# the output should be '.my_package2' instead of '{pyfile}'
importedname = "." + importedname
self._imports_stack.append((node, importedname))
@staticmethod
def _is_fallback_import(node, imports):
imports = [import_node for (import_node, _) in imports]
return any(astroid.are_exclusive(import_node, node) for import_node in imports)
def _check_imports_order(self, _module_node):
"""Checks imports of module `node` are grouped by category
Imports must follow this order: standard, 3rd party, local
"""
std_imports = []
third_party_imports = []
first_party_imports = []
# need of a list that holds third or first party ordered import
external_imports = []
local_imports = []
third_party_not_ignored = []
first_party_not_ignored = []
local_not_ignored = []
isort_obj = isort.SortImports(
file_contents="",
known_third_party=self.config.known_third_party,
known_standard_library=self.config.known_standard_library,
)
for node, modname in self._imports_stack:
if modname.startswith("."):
package = "." + modname.split(".")[1]
else:
package = modname.split(".")[0]
nested = not isinstance(node.parent, astroid.Module)
ignore_for_import_order = not self.linter.is_message_enabled(
"wrong-import-order", node.fromlineno
)
import_category = isort_obj.place_module(package)
node_and_package_import = (node, package)
if import_category in ("FUTURE", "STDLIB"):
std_imports.append(node_and_package_import)
wrong_import = (
third_party_not_ignored
or first_party_not_ignored
or local_not_ignored
)
if self._is_fallback_import(node, wrong_import):
continue
if wrong_import and not nested:
self.add_message(
"wrong-import-order",
node=node,
args=(
'standard import "%s"' % node.as_string(),
'"%s"' % wrong_import[0][0].as_string(),
),
)
elif import_category == "THIRDPARTY":
third_party_imports.append(node_and_package_import)
external_imports.append(node_and_package_import)
if not nested and not ignore_for_import_order:
third_party_not_ignored.append(node_and_package_import)
wrong_import = first_party_not_ignored or local_not_ignored
if wrong_import and not nested:
self.add_message(
"wrong-import-order",
node=node,
args=(
'third party import "%s"' % node.as_string(),
'"%s"' % wrong_import[0][0].as_string(),
),
)
elif import_category == "FIRSTPARTY":
first_party_imports.append(node_and_package_import)
external_imports.append(node_and_package_import)
if not nested and not ignore_for_import_order:
first_party_not_ignored.append(node_and_package_import)
wrong_import = local_not_ignored
if wrong_import and not nested:
self.add_message(
"wrong-import-order",
node=node,
args=(
'first party import "%s"' % node.as_string(),
'"%s"' % wrong_import[0][0].as_string(),
),
)
elif import_category == "LOCALFOLDER":
local_imports.append((node, package))
if not nested and not ignore_for_import_order:
local_not_ignored.append((node, package))
return std_imports, external_imports, local_imports
def _get_imported_module(self, importnode, modname):
try:
return importnode.do_import_module(modname)
except astroid.TooManyLevelsError:
if _ignore_import_failure(importnode, modname, self._ignored_modules):
return None
self.add_message("relative-beyond-top-level", node=importnode)
except astroid.AstroidSyntaxError as exc:
message = "Cannot import {!r} due to syntax error {!r}".format(
modname, str(exc.error) # pylint: disable=no-member; false positive
)
self.add_message("syntax-error", line=importnode.lineno, args=message)
except astroid.AstroidBuildingException:
if not self.linter.is_message_enabled("import-error"):
return None
if _ignore_import_failure(importnode, modname, self._ignored_modules):
return None
if not self.config.analyse_fallback_blocks and is_from_fallback_block(
importnode
):
return None
dotted_modname = _get_import_name(importnode, modname)
self.add_message("import-error", args=repr(dotted_modname), node=importnode)
def _check_relative_import(
self, modnode, importnode, importedmodnode, importedasname
):
"""check relative import. node is either an Import or From node, modname
the imported module name.
"""
if not self.linter.is_message_enabled("relative-import"):
return None
if importedmodnode.file is None:
return False # built-in module
if modnode is importedmodnode:
return False # module importing itself
if modnode.absolute_import_activated() or getattr(importnode, "level", None):
return False
if importedmodnode.name != importedasname:
# this must be a relative import...
self.add_message(
"relative-import",
args=(importedasname, importedmodnode.name),
node=importnode,
)
return None
return None
def _add_imported_module(self, node, importedmodname):
"""notify an imported module, used to analyze dependencies"""
module_file = node.root().file
context_name = node.root().name
base = os.path.splitext(os.path.basename(module_file))[0]
try:
importedmodname = get_module_part(importedmodname, module_file)
except ImportError:
pass
if context_name == importedmodname:
self.add_message("import-self", node=node)
elif not is_standard_module(importedmodname):
# if this is not a package __init__ module
if base != "__init__" and context_name not in self._module_pkg:
# record the module's parent, or the module itself if this is
# a top level module, as the package it belongs to
self._module_pkg[context_name] = context_name.rsplit(".", 1)[0]
# handle dependencies
importedmodnames = self.stats["dependencies"].setdefault(
importedmodname, set()
)
if context_name not in importedmodnames:
importedmodnames.add(context_name)
# update import graph
self.import_graph[context_name].add(importedmodname)
if not self.linter.is_message_enabled("cyclic-import", line=node.lineno):
self._excluded_edges[context_name].add(importedmodname)
def _check_deprecated_module(self, node, mod_path):
"""check if the module is deprecated"""
for mod_name in self.config.deprecated_modules:
if mod_path == mod_name or mod_path.startswith(mod_name + "."):
self.add_message("deprecated-module", node=node, args=mod_path)
def _check_import_as_rename(self, node):
names = node.names
for name in names:
if not all(name):
return
real_name = name[0]
splitted_packages = real_name.rsplit(".")
real_name = splitted_packages[-1]
imported_name = name[1]
# consider only following cases
# import x as x
# and ignore following
# import x.y.z as z
if real_name == imported_name and len(splitted_packages) == 1:
self.add_message("useless-import-alias", node=node)
def _check_reimport(self, node, basename=None, level=None):
"""check if the import is necessary (i.e. not already done)"""
if not self.linter.is_message_enabled("reimported"):
return
frame = node.frame()
root = node.root()
contexts = [(frame, level)]
if root is not frame:
contexts.append((root, None))
for known_context, known_level in contexts:
for name, alias in node.names:
first = _get_first_import(
node, known_context, name, basename, known_level, alias
)
if first is not None:
self.add_message(
"reimported", node=node, args=(name, first.fromlineno)
)
def _report_external_dependencies(self, sect, _, _dummy):
"""return a verbatim layout for displaying dependencies"""
dep_info = _make_tree_defs(self._external_dependencies_info().items())
if not dep_info:
raise EmptyReportError()
tree_str = _repr_tree_defs(dep_info)
sect.append(VerbatimText(tree_str))
def _report_dependencies_graph(self, sect, _, _dummy):
"""write dependencies as a dot (graphviz) file"""
dep_info = self.stats["dependencies"]
if not dep_info or not (
self.config.import_graph
or self.config.ext_import_graph
or self.config.int_import_graph
):
raise EmptyReportError()
filename = self.config.import_graph
if filename:
_make_graph(filename, dep_info, sect, "")
filename = self.config.ext_import_graph
if filename:
_make_graph(filename, self._external_dependencies_info(), sect, "external ")
filename = self.config.int_import_graph
if filename:
_make_graph(filename, self._internal_dependencies_info(), sect, "internal ")
def _filter_dependencies_graph(self, internal):
"""build the internal or the external depedency graph"""
graph = collections.defaultdict(set)
for importee, importers in self.stats["dependencies"].items():
for importer in importers:
package = self._module_pkg.get(importer, importer)
is_inside = importee.startswith(package)
if is_inside and internal or not is_inside and not internal:
graph[importee].add(importer)
return graph
@decorators.cached
def _external_dependencies_info(self):
"""return cached external dependencies information or build and
cache them
"""
return self._filter_dependencies_graph(internal=False)
@decorators.cached
def _internal_dependencies_info(self):
"""return cached internal dependencies information or build and
cache them
"""
return self._filter_dependencies_graph(internal=True)
def _check_wildcard_imports(self, node, imported_module):
if node.root().package:
# Skip the check if in __init__.py issue #2026
return
wildcard_import_is_allowed = self._wildcard_import_is_allowed(imported_module)
for name, _ in node.names:
if name == "*" and not wildcard_import_is_allowed:
self.add_message("wildcard-import", args=node.modname, node=node)
def _wildcard_import_is_allowed(self, imported_module):
return (
self.config.allow_wildcard_with_all
and imported_module is not None
and "__all__" in imported_module.locals
)
def register(linter):
"""required method to auto register this checker """
linter.register_checker(ImportsChecker(linter))

View File

@@ -0,0 +1,375 @@
# -*- coding: utf-8 -*-
# Copyright (c) 2009-2011, 2013-2014 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2009, 2012, 2014 Google, Inc.
# Copyright (c) 2012 Mike Bryant <leachim@leachim.info>
# Copyright (c) 2014 Brett Cannon <brett@python.org>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Chris Murray <chris@chrismurray.scot>
# Copyright (c) 2016 Ashley Whetter <ashley@awhetter.co.uk>
# Copyright (c) 2017 guillaume2 <guillaume.peillex@gmail.col>
# Copyright (c) 2017 Łukasz Rogalski <rogalski.91@gmail.com>
# Copyright (c) 2018 Mike Frysinger <vapier@gmail.com>
# Copyright (c) 2018 Mariatta Wijaya <mariatta@python.org>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""checker for use of Python logging
"""
import string
import astroid
from pylint import checkers
from pylint import interfaces
from pylint.checkers import utils
from pylint.checkers.utils import check_messages
MSGS = {
"W1201": (
"Specify string format arguments as logging function parameters",
"logging-not-lazy",
"Used when a logging statement has a call form of "
'"logging.<logging method>(format_string % (format_args...))". '
"Such calls should leave string interpolation to the logging "
"method itself and be written "
'"logging.<logging method>(format_string, format_args...)" '
"so that the program may avoid incurring the cost of the "
"interpolation in those cases in which no message will be "
"logged. For more, see "
"http://www.python.org/dev/peps/pep-0282/.",
),
"W1202": (
"Use % formatting in logging functions and pass the % "
"parameters as arguments",
"logging-format-interpolation",
"Used when a logging statement has a call form of "
'"logging.<logging method>(format_string.format(format_args...))"'
". Such calls should use % formatting instead, but leave "
"interpolation to the logging function by passing the parameters "
"as arguments.",
),
"W1203": (
"Use % formatting in logging functions and pass the % "
"parameters as arguments",
"logging-fstring-interpolation",
"Used when a logging statement has a call form of "
'"logging.method(f"..."))"'
". Such calls should use % formatting instead, but leave "
"interpolation to the logging function by passing the parameters "
"as arguments.",
),
"E1200": (
"Unsupported logging format character %r (%#02x) at index %d",
"logging-unsupported-format",
"Used when an unsupported format character is used in a logging "
"statement format string.",
),
"E1201": (
"Logging format string ends in middle of conversion specifier",
"logging-format-truncated",
"Used when a logging statement format string terminates before "
"the end of a conversion specifier.",
),
"E1205": (
"Too many arguments for logging format string",
"logging-too-many-args",
"Used when a logging format string is given too many arguments.",
),
"E1206": (
"Not enough arguments for logging format string",
"logging-too-few-args",
"Used when a logging format string is given too few arguments.",
),
}
CHECKED_CONVENIENCE_FUNCTIONS = {
"critical",
"debug",
"error",
"exception",
"fatal",
"info",
"warn",
"warning",
}
def is_method_call(func, types=(), methods=()):
"""Determines if a BoundMethod node represents a method call.
Args:
func (astroid.BoundMethod): The BoundMethod AST node to check.
types (Optional[String]): Optional sequence of caller type names to restrict check.
methods (Optional[String]): Optional sequence of method names to restrict check.
Returns:
bool: true if the node represents a method call for the given type and
method names, False otherwise.
"""
return (
isinstance(func, astroid.BoundMethod)
and isinstance(func.bound, astroid.Instance)
and (func.bound.name in types if types else True)
and (func.name in methods if methods else True)
)
class LoggingChecker(checkers.BaseChecker):
"""Checks use of the logging module."""
__implements__ = interfaces.IAstroidChecker
name = "logging"
msgs = MSGS
options = (
(
"logging-modules",
{
"default": ("logging",),
"type": "csv",
"metavar": "<comma separated list>",
"help": "Logging modules to check that the string format "
"arguments are in logging function parameter format.",
},
),
(
"logging-format-style",
{
"default": "old",
"type": "choice",
"metavar": "<old (%) or new ({)>",
"choices": ["old", "new"],
"help": "Format style used to check logging format string. "
"`old` means using % formatting, while `new` is for `{}` formatting.",
},
),
)
def visit_module(self, node): # pylint: disable=unused-argument
"""Clears any state left in this checker from last module checked."""
# The code being checked can just as easily "import logging as foo",
# so it is necessary to process the imports and store in this field
# what name the logging module is actually given.
self._logging_names = set()
logging_mods = self.config.logging_modules
self._format_style = self.config.logging_format_style
self._logging_modules = set(logging_mods)
self._from_imports = {}
for logging_mod in logging_mods:
parts = logging_mod.rsplit(".", 1)
if len(parts) > 1:
self._from_imports[parts[0]] = parts[1]
def visit_importfrom(self, node):
"""Checks to see if a module uses a non-Python logging module."""
try:
logging_name = self._from_imports[node.modname]
for module, as_name in node.names:
if module == logging_name:
self._logging_names.add(as_name or module)
except KeyError:
pass
def visit_import(self, node):
"""Checks to see if this module uses Python's built-in logging."""
for module, as_name in node.names:
if module in self._logging_modules:
self._logging_names.add(as_name or module)
@check_messages(*MSGS)
def visit_call(self, node):
"""Checks calls to logging methods."""
def is_logging_name():
return (
isinstance(node.func, astroid.Attribute)
and isinstance(node.func.expr, astroid.Name)
and node.func.expr.name in self._logging_names
)
def is_logger_class():
try:
for inferred in node.func.infer():
if isinstance(inferred, astroid.BoundMethod):
parent = inferred._proxied.parent
if isinstance(parent, astroid.ClassDef) and (
parent.qname() == "logging.Logger"
or any(
ancestor.qname() == "logging.Logger"
for ancestor in parent.ancestors()
)
):
return True, inferred._proxied.name
except astroid.exceptions.InferenceError:
pass
return False, None
if is_logging_name():
name = node.func.attrname
else:
result, name = is_logger_class()
if not result:
return
self._check_log_method(node, name)
def _check_log_method(self, node, name):
"""Checks calls to logging.log(level, format, *format_args)."""
if name == "log":
if node.starargs or node.kwargs or len(node.args) < 2:
# Either a malformed call, star args, or double-star args. Beyond
# the scope of this checker.
return
format_pos = 1
elif name in CHECKED_CONVENIENCE_FUNCTIONS:
if node.starargs or node.kwargs or not node.args:
# Either no args, star args, or double-star args. Beyond the
# scope of this checker.
return
format_pos = 0
else:
return
if isinstance(node.args[format_pos], astroid.BinOp):
binop = node.args[format_pos]
emit = binop.op == "%"
if binop.op == "+":
total_number_of_strings = sum(
1
for operand in (binop.left, binop.right)
if self._is_operand_literal_str(utils.safe_infer(operand))
)
emit = total_number_of_strings > 0
if emit:
self.add_message("logging-not-lazy", node=node)
elif isinstance(node.args[format_pos], astroid.Call):
self._check_call_func(node.args[format_pos])
elif isinstance(node.args[format_pos], astroid.Const):
self._check_format_string(node, format_pos)
elif isinstance(
node.args[format_pos], (astroid.FormattedValue, astroid.JoinedStr)
):
self.add_message("logging-fstring-interpolation", node=node)
@staticmethod
def _is_operand_literal_str(operand):
"""
Return True if the operand in argument is a literal string
"""
return isinstance(operand, astroid.Const) and operand.name == "str"
def _check_call_func(self, node):
"""Checks that function call is not format_string.format().
Args:
node (astroid.node_classes.Call):
Call AST node to be checked.
"""
func = utils.safe_infer(node.func)
types = ("str", "unicode")
methods = ("format",)
if is_method_call(func, types, methods) and not is_complex_format_str(
func.bound
):
self.add_message("logging-format-interpolation", node=node)
def _check_format_string(self, node, format_arg):
"""Checks that format string tokens match the supplied arguments.
Args:
node (astroid.node_classes.NodeNG): AST node to be checked.
format_arg (int): Index of the format string in the node arguments.
"""
num_args = _count_supplied_tokens(node.args[format_arg + 1 :])
if not num_args:
# If no args were supplied the string is not interpolated and can contain
# formatting characters - it's used verbatim. Don't check any further.
return
format_string = node.args[format_arg].value
if not isinstance(format_string, str):
# If the log format is constant non-string (e.g. logging.debug(5)),
# ensure there are no arguments.
required_num_args = 0
else:
try:
if self._format_style == "old":
keyword_args, required_num_args, _, _ = utils.parse_format_string(
format_string
)
if keyword_args:
# Keyword checking on logging strings is complicated by
# special keywords - out of scope.
return
elif self._format_style == "new":
keyword_arguments, implicit_pos_args, explicit_pos_args = utils.parse_format_method_string(
format_string
)
keyword_args_cnt = len(
set(k for k, l in keyword_arguments if not isinstance(k, int))
)
required_num_args = (
keyword_args_cnt + implicit_pos_args + explicit_pos_args
)
except utils.UnsupportedFormatCharacter as ex:
char = format_string[ex.index]
self.add_message(
"logging-unsupported-format",
node=node,
args=(char, ord(char), ex.index),
)
return
except utils.IncompleteFormatString:
self.add_message("logging-format-truncated", node=node)
return
if num_args > required_num_args:
self.add_message("logging-too-many-args", node=node)
elif num_args < required_num_args:
self.add_message("logging-too-few-args", node=node)
def is_complex_format_str(node):
"""Checks if node represents a string with complex formatting specs.
Args:
node (astroid.node_classes.NodeNG): AST node to check
Returns:
bool: True if inferred string uses complex formatting, False otherwise
"""
inferred = utils.safe_infer(node)
if inferred is None or not isinstance(inferred.value, str):
return True
try:
parsed = list(string.Formatter().parse(inferred.value))
except ValueError:
# This format string is invalid
return False
for _, _, format_spec, _ in parsed:
if format_spec:
return True
return False
def _count_supplied_tokens(args):
"""Counts the number of tokens in an args list.
The Python log functions allow for special keyword arguments: func,
exc_info and extra. To handle these cases correctly, we only count
arguments that aren't keywords.
Args:
args (list): AST nodes that are arguments for a log format string.
Returns:
int: Number of AST nodes that aren't keywords.
"""
return sum(1 for arg in args if not isinstance(arg, astroid.Keyword))
def register(linter):
"""Required method to auto-register this checker."""
linter.register_checker(LoggingChecker(linter))

View File

@@ -0,0 +1,182 @@
# -*- coding: utf-8 -*-
# Copyright (c) 2006, 2009-2013 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2012-2014 Google, Inc.
# Copyright (c) 2014-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Brett Cannon <brett@python.org>
# Copyright (c) 2014 Alexandru Coman <fcoman@bitdefender.com>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Łukasz Rogalski <rogalski.91@gmail.com>
# Copyright (c) 2016 glegoux <gilles.legoux@gmail.com>
# Copyright (c) 2017-2018 hippo91 <guillaume.peillex@gmail.com>
# Copyright (c) 2017 Mikhail Fesenko <proggga@gmail.com>
# Copyright (c) 2018 Ville Skyttä <ville.skytta@iki.fi>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""Check source code is ascii only or has an encoding declaration (PEP 263)"""
# pylint: disable=W0511
import tokenize
import re
from pylint.interfaces import IRawChecker, ITokenChecker
from pylint.checkers import BaseChecker
from pylint.utils import OPTION_RGX, MessagesHandlerMixIn
class ByIdManagedMessagesChecker(BaseChecker):
"""checks for messages that are enabled or disabled by id instead of symbol."""
__implements__ = IRawChecker
# configuration section name
name = "miscellaneous"
msgs = {
"I0023": (
"%s",
"use-symbolic-message-instead",
"Used when a message is enabled or disabled by id.",
)
}
options = ()
def process_module(self, module):
"""inspect the source file to find messages activated or deactivated by id."""
managed_msgs = MessagesHandlerMixIn.get_by_id_managed_msgs()
for (mod_name, msg_id, msg_symbol, lineno, is_disabled) in managed_msgs:
if mod_name == module.name:
if is_disabled:
txt = "Id '{ident}' is used to disable '{symbol}' message emission".format(
ident=msg_id, symbol=msg_symbol
)
else:
txt = "Id '{ident}' is used to enable '{symbol}' message emission".format(
ident=msg_id, symbol=msg_symbol
)
self.add_message("use-symbolic-message-instead", line=lineno, args=txt)
MessagesHandlerMixIn.clear_by_id_managed_msgs()
class EncodingChecker(BaseChecker):
"""checks for:
* warning notes in the code like FIXME, XXX
* encoding issues.
"""
__implements__ = (IRawChecker, ITokenChecker)
# configuration section name
name = "miscellaneous"
msgs = {
"W0511": (
"%s",
"fixme",
"Used when a warning note as FIXME or XXX is detected.",
),
"W0512": (
'Cannot decode using encoding "%s", unexpected byte at position %d',
"invalid-encoded-data",
"Used when a source line cannot be decoded using the specified "
"source file encoding.",
{"maxversion": (3, 0)},
),
}
options = (
(
"notes",
{
"type": "csv",
"metavar": "<comma separated values>",
"default": ("FIXME", "XXX", "TODO"),
"help": (
"List of note tags to take in consideration, "
"separated by a comma."
),
},
),
)
def open(self):
super().open()
self._fixme_pattern = re.compile(
r"#\s*(%s)\b" % "|".join(map(re.escape, self.config.notes)), re.I
)
def _check_encoding(self, lineno, line, file_encoding):
try:
return line.decode(file_encoding)
except UnicodeDecodeError as ex:
self.add_message(
"invalid-encoded-data", line=lineno, args=(file_encoding, ex.args[2])
)
except LookupError as ex:
if line.startswith("#") and "coding" in line and file_encoding in line:
self.add_message(
"syntax-error",
line=lineno,
args='Cannot decode using encoding "{}",'
" bad encoding".format(file_encoding),
)
def process_module(self, module):
"""inspect the source file to find encoding problem"""
if module.file_encoding:
encoding = module.file_encoding
else:
encoding = "ascii"
with module.stream() as stream:
for lineno, line in enumerate(stream):
self._check_encoding(lineno + 1, line, encoding)
def process_tokens(self, tokens):
"""inspect the source to find fixme problems"""
if not self.config.notes:
return
comments = (
token_info for token_info in tokens if token_info.type == tokenize.COMMENT
)
for comment in comments:
comment_text = comment.string[1:].lstrip() # trim '#' and whitespaces
# handle pylint disable clauses
disable_option_match = OPTION_RGX.search(comment_text)
if disable_option_match:
try:
_, value = disable_option_match.group(1).split("=", 1)
values = [_val.strip().upper() for _val in value.split(",")]
if set(values) & set(self.config.notes):
continue
except ValueError:
self.add_message(
"bad-inline-option",
args=disable_option_match.group(1).strip(),
line=comment.string,
)
continue
# emit warnings if necessary
match = self._fixme_pattern.search("#" + comment_text.lower())
if match:
note = match.group(1)
self.add_message(
"fixme",
col_offset=comment.string.lower().index(note.lower()),
args=comment_text,
line=comment.start[0],
)
def register(linter):
"""required method to auto register this checker"""
linter.register_checker(EncodingChecker(linter))
linter.register_checker(ByIdManagedMessagesChecker(linter))

View File

@@ -0,0 +1,142 @@
# Copyright (c) 2006, 2008-2011, 2013-2014 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2012-2014 Google, Inc.
# Copyright (c) 2013-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Michal Nowikowski <godfryd@gmail.com>
# Copyright (c) 2014 Brett Cannon <brett@python.org>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Alexander Todorov <atodorov@otb.bg>
# Copyright (c) 2016 Jakub Wilk <jwilk@jwilk.net>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""check for new / old style related problems
"""
import sys
import astroid
from pylint.interfaces import IAstroidChecker
from pylint.checkers import BaseChecker
from pylint.checkers.utils import check_messages, node_frame_class, has_known_bases
MSGS = {
"E1003": (
"Bad first argument %r given to super()",
"bad-super-call",
"Used when another argument than the current class is given as "
"first argument of the super builtin.",
),
"E1004": (
"Missing argument to super()",
"missing-super-argument",
"Used when the super builtin didn't receive an argument.",
{"maxversion": (3, 0)},
),
}
class NewStyleConflictChecker(BaseChecker):
"""checks for usage of new style capabilities on old style classes and
other new/old styles conflicts problems
* use of property, __slots__, super
* "super" usage
"""
__implements__ = (IAstroidChecker,)
# configuration section name
name = "newstyle"
# messages
msgs = MSGS
priority = -2
# configuration options
options = ()
@check_messages("bad-super-call", "missing-super-argument")
def visit_functiondef(self, node):
"""check use of super"""
# ignore actual functions or method within a new style class
if not node.is_method():
return
klass = node.parent.frame()
for stmt in node.nodes_of_class(astroid.Call):
if node_frame_class(stmt) != node_frame_class(node):
# Don't look down in other scopes.
continue
expr = stmt.func
if not isinstance(expr, astroid.Attribute):
continue
call = expr.expr
# skip the test if using super
if not (
isinstance(call, astroid.Call)
and isinstance(call.func, astroid.Name)
and call.func.name == "super"
):
continue
if not klass.newstyle and has_known_bases(klass):
# super should not be used on an old style class
continue
else:
# super first arg should be the class
if not call.args:
if sys.version_info[0] == 3:
# unless Python 3
continue
else:
self.add_message("missing-super-argument", node=call)
continue
# calling super(type(self), self) can lead to recursion loop
# in derived classes
arg0 = call.args[0]
if (
isinstance(arg0, astroid.Call)
and isinstance(arg0.func, astroid.Name)
and arg0.func.name == "type"
):
self.add_message("bad-super-call", node=call, args=("type",))
continue
# calling super(self.__class__, self) can lead to recursion loop
# in derived classes
if (
len(call.args) >= 2
and isinstance(call.args[1], astroid.Name)
and call.args[1].name == "self"
and isinstance(arg0, astroid.Attribute)
and arg0.attrname == "__class__"
):
self.add_message(
"bad-super-call", node=call, args=("self.__class__",)
)
continue
try:
supcls = call.args and next(call.args[0].infer(), None)
except astroid.InferenceError:
continue
if klass is not supcls:
name = None
# if supcls is not Uninferable, then supcls was infered
# and use its name. Otherwise, try to look
# for call.args[0].name
if supcls:
name = supcls.name
elif call.args and hasattr(call.args[0], "name"):
name = call.args[0].name
if name:
self.add_message("bad-super-call", node=call, args=(name,))
visit_asyncfunctiondef = visit_functiondef
def register(linter):
"""required method to auto register this checker """
linter.register_checker(NewStyleConflictChecker(linter))

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,125 @@
# Copyright (c) 2007, 2010, 2013, 2015 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2013 Google, Inc.
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015-2017 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2015 Mike Frysinger <vapier@gentoo.org>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Glenn Matthews <glenn@e-dad.net>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
""" Copyright (c) 2003-2010 LOGILAB S.A. (Paris, FRANCE).
http://www.logilab.fr/ -- mailto:contact@logilab.fr
Raw metrics checker
"""
import tokenize
from typing import Any
from pylint.interfaces import ITokenChecker
from pylint.exceptions import EmptyReportError
from pylint.checkers import BaseTokenChecker
from pylint.reporters import diff_string
from pylint.reporters.ureports.nodes import Table
def report_raw_stats(sect, stats, old_stats):
"""calculate percentage of code / doc / comment / empty
"""
total_lines = stats["total_lines"]
if not total_lines:
raise EmptyReportError()
sect.description = "%s lines have been analyzed" % total_lines
lines = ("type", "number", "%", "previous", "difference")
for node_type in ("code", "docstring", "comment", "empty"):
key = node_type + "_lines"
total = stats[key]
percent = float(total * 100) / total_lines
old = old_stats.get(key, None)
if old is not None:
diff_str = diff_string(old, total)
else:
old, diff_str = "NC", "NC"
lines += (node_type, str(total), "%.2f" % percent, str(old), diff_str)
sect.append(Table(children=lines, cols=5, rheaders=1))
class RawMetricsChecker(BaseTokenChecker):
"""does not check anything but gives some raw metrics :
* total number of lines
* total number of code lines
* total number of docstring lines
* total number of comments lines
* total number of empty lines
"""
__implements__ = (ITokenChecker,)
# configuration section name
name = "metrics"
# configuration options
options = ()
# messages
msgs = {} # type: Any
# reports
reports = (("RP0701", "Raw metrics", report_raw_stats),)
def __init__(self, linter):
BaseTokenChecker.__init__(self, linter)
self.stats = None
def open(self):
"""init statistics"""
self.stats = self.linter.add_stats(
total_lines=0,
code_lines=0,
empty_lines=0,
docstring_lines=0,
comment_lines=0,
)
def process_tokens(self, tokens):
"""update stats"""
i = 0
tokens = list(tokens)
while i < len(tokens):
i, lines_number, line_type = get_type(tokens, i)
self.stats["total_lines"] += lines_number
self.stats[line_type] += lines_number
JUNK = (tokenize.NL, tokenize.INDENT, tokenize.NEWLINE, tokenize.ENDMARKER)
def get_type(tokens, start_index):
"""return the line type : docstring, comment, code, empty"""
i = start_index
tok_type = tokens[i][0]
start = tokens[i][2]
pos = start
line_type = None
while i < len(tokens) and tokens[i][2][0] == start[0]:
tok_type = tokens[i][0]
pos = tokens[i][3]
if line_type is None:
if tok_type == tokenize.STRING:
line_type = "docstring_lines"
elif tok_type == tokenize.COMMENT:
line_type = "comment_lines"
elif tok_type in JUNK:
pass
else:
line_type = "code_lines"
i += 1
if line_type is None:
line_type = "empty_lines"
elif i < len(tokens) and tokens[i][0] == tokenize.NEWLINE:
i += 1
return i, pos[0] - start[0] + 1, line_type
def register(linter):
""" required method to auto register this checker """
linter.register_checker(RawMetricsChecker(linter))

File diff suppressed because it is too large Load Diff

View File

@@ -0,0 +1,449 @@
# Copyright (c) 2006, 2008-2014 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2012 Ry4an Brase <ry4an-hg@ry4an.org>
# Copyright (c) 2012 Google, Inc.
# Copyright (c) 2012 Anthony VEREZ <anthony.verez.external@cassidian.com>
# Copyright (c) 2014-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Brett Cannon <brett@python.org>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2017 Anthony Sottile <asottile@umich.edu>
# Copyright (c) 2017 Mikhail Fesenko <proggga@gmail.com>
# Copyright (c) 2018 ssolanki <sushobhitsolanki@gmail.com>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
# pylint: disable=W0622
"""a similarities / code duplication command line tool and pylint checker
"""
from __future__ import print_function
import sys
from collections import defaultdict
from itertools import groupby
import astroid
from pylint.utils import decoding_stream
from pylint.interfaces import IRawChecker
from pylint.checkers import BaseChecker, table_lines_from_stats
from pylint.reporters.ureports.nodes import Table
class Similar:
"""finds copy-pasted lines of code in a project"""
def __init__(
self,
min_lines=4,
ignore_comments=False,
ignore_docstrings=False,
ignore_imports=False,
):
self.min_lines = min_lines
self.ignore_comments = ignore_comments
self.ignore_docstrings = ignore_docstrings
self.ignore_imports = ignore_imports
self.linesets = []
def append_stream(self, streamid, stream, encoding=None):
"""append a file to search for similarities"""
if encoding is None:
readlines = stream.readlines
else:
readlines = decoding_stream(stream, encoding).readlines
try:
self.linesets.append(
LineSet(
streamid,
readlines(),
self.ignore_comments,
self.ignore_docstrings,
self.ignore_imports,
)
)
except UnicodeDecodeError:
pass
def run(self):
"""start looking for similarities and display results on stdout"""
self._display_sims(self._compute_sims())
def _compute_sims(self):
"""compute similarities in appended files"""
no_duplicates = defaultdict(list)
for num, lineset1, idx1, lineset2, idx2 in self._iter_sims():
duplicate = no_duplicates[num]
for couples in duplicate:
if (lineset1, idx1) in couples or (lineset2, idx2) in couples:
couples.add((lineset1, idx1))
couples.add((lineset2, idx2))
break
else:
duplicate.append({(lineset1, idx1), (lineset2, idx2)})
sims = []
for num, ensembles in no_duplicates.items():
for couples in ensembles:
sims.append((num, couples))
sims.sort()
sims.reverse()
return sims
def _display_sims(self, sims):
"""display computed similarities on stdout"""
nb_lignes_dupliquees = 0
for num, couples in sims:
print()
print(num, "similar lines in", len(couples), "files")
couples = sorted(couples)
for lineset, idx in couples:
print("==%s:%s" % (lineset.name, idx))
# pylint: disable=W0631
for line in lineset._real_lines[idx : idx + num]:
print(" ", line.rstrip())
nb_lignes_dupliquees += num * (len(couples) - 1)
nb_total_lignes = sum([len(lineset) for lineset in self.linesets])
print(
"TOTAL lines=%s duplicates=%s percent=%.2f"
% (
nb_total_lignes,
nb_lignes_dupliquees,
nb_lignes_dupliquees * 100. / nb_total_lignes,
)
)
def _find_common(self, lineset1, lineset2):
"""find similarities in the two given linesets"""
lines1 = lineset1.enumerate_stripped
lines2 = lineset2.enumerate_stripped
find = lineset2.find
index1 = 0
min_lines = self.min_lines
while index1 < len(lineset1):
skip = 1
num = 0
for index2 in find(lineset1[index1]):
non_blank = 0
for num, ((_, line1), (_, line2)) in enumerate(
zip(lines1(index1), lines2(index2))
):
if line1 != line2:
if non_blank > min_lines:
yield num, lineset1, index1, lineset2, index2
skip = max(skip, num)
break
if line1:
non_blank += 1
else:
# we may have reach the end
num += 1
if non_blank > min_lines:
yield num, lineset1, index1, lineset2, index2
skip = max(skip, num)
index1 += skip
def _iter_sims(self):
"""iterate on similarities among all files, by making a cartesian
product
"""
for idx, lineset in enumerate(self.linesets[:-1]):
for lineset2 in self.linesets[idx + 1 :]:
for sim in self._find_common(lineset, lineset2):
yield sim
def stripped_lines(lines, ignore_comments, ignore_docstrings, ignore_imports):
"""return lines with leading/trailing whitespace and any ignored code
features removed
"""
if ignore_imports:
tree = astroid.parse("".join(lines))
node_is_import_by_lineno = (
(node.lineno, isinstance(node, (astroid.Import, astroid.ImportFrom)))
for node in tree.body
)
line_begins_import = {
lineno: all(is_import for _, is_import in node_is_import_group)
for lineno, node_is_import_group in groupby(
node_is_import_by_lineno, key=lambda x: x[0]
)
}
current_line_is_import = False
strippedlines = []
docstring = None
for lineno, line in enumerate(lines, start=1):
line = line.strip()
if ignore_docstrings:
if not docstring and (line.startswith('"""') or line.startswith("'''")):
docstring = line[:3]
line = line[3:]
if docstring:
if line.endswith(docstring):
docstring = None
line = ""
if ignore_imports:
current_line_is_import = line_begins_import.get(
lineno, current_line_is_import
)
if current_line_is_import:
line = ""
if ignore_comments:
# XXX should use regex in checkers/format to avoid cutting
# at a "#" in a string
line = line.split("#", 1)[0].strip()
strippedlines.append(line)
return strippedlines
class LineSet:
"""Holds and indexes all the lines of a single source file"""
def __init__(
self,
name,
lines,
ignore_comments=False,
ignore_docstrings=False,
ignore_imports=False,
):
self.name = name
self._real_lines = lines
self._stripped_lines = stripped_lines(
lines, ignore_comments, ignore_docstrings, ignore_imports
)
self._index = self._mk_index()
def __str__(self):
return "<Lineset for %s>" % self.name
def __len__(self):
return len(self._real_lines)
def __getitem__(self, index):
return self._stripped_lines[index]
def __lt__(self, other):
return self.name < other.name
def __hash__(self):
return id(self)
def enumerate_stripped(self, start_at=0):
"""return an iterator on stripped lines, starting from a given index
if specified, else 0
"""
idx = start_at
if start_at:
lines = self._stripped_lines[start_at:]
else:
lines = self._stripped_lines
for line in lines:
# if line:
yield idx, line
idx += 1
def find(self, stripped_line):
"""return positions of the given stripped line in this set"""
return self._index.get(stripped_line, ())
def _mk_index(self):
"""create the index for this set"""
index = defaultdict(list)
for line_no, line in enumerate(self._stripped_lines):
if line:
index[line].append(line_no)
return index
MSGS = {
"R0801": (
"Similar lines in %s files\n%s",
"duplicate-code",
"Indicates that a set of similar lines has been detected "
"among multiple file. This usually means that the code should "
"be refactored to avoid this duplication.",
)
}
def report_similarities(sect, stats, old_stats):
"""make a layout with some stats about duplication"""
lines = ["", "now", "previous", "difference"]
lines += table_lines_from_stats(
stats, old_stats, ("nb_duplicated_lines", "percent_duplicated_lines")
)
sect.append(Table(children=lines, cols=4, rheaders=1, cheaders=1))
# wrapper to get a pylint checker from the similar class
class SimilarChecker(BaseChecker, Similar):
"""checks for similarities and duplicated code. This computation may be
memory / CPU intensive, so you should disable it if you experiment some
problems.
"""
__implements__ = (IRawChecker,)
# configuration section name
name = "similarities"
# messages
msgs = MSGS
# configuration options
# for available dict keys/values see the optik parser 'add_option' method
options = (
(
"min-similarity-lines", # type: ignore
{
"default": 4,
"type": "int",
"metavar": "<int>",
"help": "Minimum lines number of a similarity.",
},
),
(
"ignore-comments",
{
"default": True,
"type": "yn",
"metavar": "<y or n>",
"help": "Ignore comments when computing similarities.",
},
),
(
"ignore-docstrings",
{
"default": True,
"type": "yn",
"metavar": "<y or n>",
"help": "Ignore docstrings when computing similarities.",
},
),
(
"ignore-imports",
{
"default": False,
"type": "yn",
"metavar": "<y or n>",
"help": "Ignore imports when computing similarities.",
},
),
)
# reports
reports = (("RP0801", "Duplication", report_similarities),) # type: ignore
def __init__(self, linter=None):
BaseChecker.__init__(self, linter)
Similar.__init__(
self, min_lines=4, ignore_comments=True, ignore_docstrings=True
)
self.stats = None
def set_option(self, optname, value, action=None, optdict=None):
"""method called to set an option (registered in the options list)
overridden to report options setting to Similar
"""
BaseChecker.set_option(self, optname, value, action, optdict)
if optname == "min-similarity-lines":
self.min_lines = self.config.min_similarity_lines
elif optname == "ignore-comments":
self.ignore_comments = self.config.ignore_comments
elif optname == "ignore-docstrings":
self.ignore_docstrings = self.config.ignore_docstrings
elif optname == "ignore-imports":
self.ignore_imports = self.config.ignore_imports
def open(self):
"""init the checkers: reset linesets and statistics information"""
self.linesets = []
self.stats = self.linter.add_stats(
nb_duplicated_lines=0, percent_duplicated_lines=0
)
def process_module(self, node):
"""process a module
the module's content is accessible via the stream object
stream must implement the readlines method
"""
with node.stream() as stream:
self.append_stream(self.linter.current_name, stream, node.file_encoding)
def close(self):
"""compute and display similarities on closing (i.e. end of parsing)"""
total = sum(len(lineset) for lineset in self.linesets)
duplicated = 0
stats = self.stats
for num, couples in self._compute_sims():
msg = []
for lineset, idx in couples:
msg.append("==%s:%s" % (lineset.name, idx))
msg.sort()
# pylint: disable=W0631
for line in lineset._real_lines[idx : idx + num]:
msg.append(line.rstrip())
self.add_message("R0801", args=(len(couples), "\n".join(msg)))
duplicated += num * (len(couples) - 1)
stats["nb_duplicated_lines"] = duplicated
stats["percent_duplicated_lines"] = total and duplicated * 100. / total
def register(linter):
"""required method to auto register this checker """
linter.register_checker(SimilarChecker(linter))
def usage(status=0):
"""display command line usage information"""
print("finds copy pasted blocks in a set of files")
print()
print(
"Usage: symilar [-d|--duplicates min_duplicated_lines] \
[-i|--ignore-comments] [--ignore-docstrings] [--ignore-imports] file1..."
)
sys.exit(status)
def Run(argv=None):
"""standalone command line access point"""
if argv is None:
argv = sys.argv[1:]
from getopt import getopt
s_opts = "hdi"
l_opts = (
"help",
"duplicates=",
"ignore-comments",
"ignore-imports",
"ignore-docstrings",
)
min_lines = 4
ignore_comments = False
ignore_docstrings = False
ignore_imports = False
opts, args = getopt(argv, s_opts, l_opts)
for opt, val in opts:
if opt in ("-d", "--duplicates"):
min_lines = int(val)
elif opt in ("-h", "--help"):
usage()
elif opt in ("-i", "--ignore-comments"):
ignore_comments = True
elif opt in ("--ignore-docstrings",):
ignore_docstrings = True
elif opt in ("--ignore-imports",):
ignore_imports = True
if not args:
usage(1)
sim = Similar(min_lines, ignore_comments, ignore_docstrings, ignore_imports)
for filename in args:
with open(filename) as stream:
sim.append_stream(filename, stream)
sim.run()
sys.exit(0)
if __name__ == "__main__":
Run()

View File

@@ -0,0 +1,403 @@
# -*- coding: utf-8 -*-
# Copyright (c) 2014-2017 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Michal Nowikowski <godfryd@gmail.com>
# Copyright (c) 2014 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2015 Pavel Roskin <proski@gnu.org>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016-2017 Pedro Algarvio <pedro@algarvio.me>
# Copyright (c) 2016 Alexander Todorov <atodorov@otb.bg>
# Copyright (c) 2017 Łukasz Rogalski <rogalski.91@gmail.com>
# Copyright (c) 2017 Mikhail Fesenko <proggga@gmail.com>
# Copyright (c) 2018 Mike Frysinger <vapier@gmail.com>
# Copyright (c) 2018 Sushobhit <31987769+sushobhit27@users.noreply.github.com>
# Copyright (c) 2018 Anthony Sottile <asottile@umich.edu>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""Checker for spelling errors in comments and docstrings.
"""
import os
import tokenize
import re
try:
import enchant
from enchant.tokenize import ( # type: ignore
get_tokenizer,
Chunker,
Filter,
EmailFilter,
URLFilter,
WikiWordFilter,
)
except ImportError:
enchant = None
# pylint: disable=no-init
class Filter: # type: ignore
def _skip(self, word):
raise NotImplementedError
class Chunker: # type: ignore
pass
from pylint.interfaces import ITokenChecker, IAstroidChecker
from pylint.checkers import BaseTokenChecker
from pylint.checkers.utils import check_messages
if enchant is not None:
br = enchant.Broker()
dicts = br.list_dicts()
dict_choices = [""] + [d[0] for d in dicts]
dicts = ["%s (%s)" % (d[0], d[1].name) for d in dicts]
dicts = ", ".join(dicts)
instr = ""
else:
dicts = "none"
dict_choices = [""]
instr = " To make it working install python-enchant package."
class WordsWithDigigtsFilter(Filter):
"""Skips words with digits.
"""
def _skip(self, word):
for char in word:
if char.isdigit():
return True
return False
class WordsWithUnderscores(Filter):
"""Skips words with underscores.
They are probably function parameter names.
"""
def _skip(self, word):
return "_" in word
class CamelCasedWord(Filter):
r"""Filter skipping over camelCasedWords.
This filter skips any words matching the following regular expression:
^([a-z]\w+[A-Z]+\w+)
That is, any words that are camelCasedWords.
"""
_pattern = re.compile(r"^([a-z]+([\d]|[A-Z])(?:\w+)?)")
def _skip(self, word):
return bool(self._pattern.match(word))
class SphinxDirectives(Filter):
r"""Filter skipping over Sphinx Directives.
This filter skips any words matching the following regular expression:
^:([a-z]+):`([^`]+)(`)?
That is, for example, :class:`BaseQuery`
"""
# The final ` in the pattern is optional because enchant strips it out
_pattern = re.compile(r"^:([a-z]+):`([^`]+)(`)?")
def _skip(self, word):
return bool(self._pattern.match(word))
class ForwardSlashChunkder(Chunker):
"""
This chunker allows splitting words like 'before/after' into 'before' and 'after'
"""
def next(self):
while True:
if not self._text:
raise StopIteration()
if "/" not in self._text:
text = self._text
self._offset = 0
self._text = ""
return (text, 0)
pre_text, post_text = self._text.split("/", 1)
self._text = post_text
self._offset = 0
if (
not pre_text
or not post_text
or not pre_text[-1].isalpha()
or not post_text[0].isalpha()
):
self._text = ""
self._offset = 0
return (pre_text + "/" + post_text, 0)
return (pre_text, 0)
def _next(self):
while True:
if "/" not in self._text:
return (self._text, 0)
pre_text, post_text = self._text.split("/", 1)
if not pre_text or not post_text:
break
if not pre_text[-1].isalpha() or not post_text[0].isalpha():
raise StopIteration()
self._text = pre_text + " " + post_text
raise StopIteration()
class SpellingChecker(BaseTokenChecker):
"""Check spelling in comments and docstrings"""
__implements__ = (ITokenChecker, IAstroidChecker)
name = "spelling"
msgs = {
"C0401": (
"Wrong spelling of a word '%s' in a comment:\n%s\n"
"%s\nDid you mean: '%s'?",
"wrong-spelling-in-comment",
"Used when a word in comment is not spelled correctly.",
),
"C0402": (
"Wrong spelling of a word '%s' in a docstring:\n%s\n"
"%s\nDid you mean: '%s'?",
"wrong-spelling-in-docstring",
"Used when a word in docstring is not spelled correctly.",
),
"C0403": (
"Invalid characters %r in a docstring",
"invalid-characters-in-docstring",
"Used when a word in docstring cannot be checked by enchant.",
),
}
options = (
(
"spelling-dict",
{
"default": "",
"type": "choice",
"metavar": "<dict name>",
"choices": dict_choices,
"help": "Spelling dictionary name. "
"Available dictionaries: %s.%s." % (dicts, instr),
},
),
(
"spelling-ignore-words",
{
"default": "",
"type": "string",
"metavar": "<comma separated words>",
"help": "List of comma separated words that " "should not be checked.",
},
),
(
"spelling-private-dict-file",
{
"default": "",
"type": "string",
"metavar": "<path to file>",
"help": "A path to a file that contains private "
"dictionary; one word per line.",
},
),
(
"spelling-store-unknown-words",
{
"default": "n",
"type": "yn",
"metavar": "<y_or_n>",
"help": "Tells whether to store unknown words to "
"indicated private dictionary in "
"--spelling-private-dict-file option instead of "
"raising a message.",
},
),
(
"max-spelling-suggestions",
{
"default": 4,
"type": "int",
"metavar": "N",
"help": "Limits count of emitted suggestions for " "spelling mistakes.",
},
),
)
def open(self):
self.initialized = False
self.private_dict_file = None
if enchant is None:
return
dict_name = self.config.spelling_dict
if not dict_name:
return
self.ignore_list = [
w.strip() for w in self.config.spelling_ignore_words.split(",")
]
# "param" appears in docstring in param description and
# "pylint" appears in comments in pylint pragmas.
self.ignore_list.extend(["param", "pylint"])
# Expand tilde to allow e.g. spelling-private-dict-file = ~/.pylintdict
if self.config.spelling_private_dict_file:
self.config.spelling_private_dict_file = os.path.expanduser(
self.config.spelling_private_dict_file
)
if self.config.spelling_private_dict_file:
self.spelling_dict = enchant.DictWithPWL(
dict_name, self.config.spelling_private_dict_file
)
self.private_dict_file = open(self.config.spelling_private_dict_file, "a")
else:
self.spelling_dict = enchant.Dict(dict_name)
if self.config.spelling_store_unknown_words:
self.unknown_words = set()
self.tokenizer = get_tokenizer(
dict_name,
chunkers=[ForwardSlashChunkder],
filters=[
EmailFilter,
URLFilter,
WikiWordFilter,
WordsWithDigigtsFilter,
WordsWithUnderscores,
CamelCasedWord,
SphinxDirectives,
],
)
self.initialized = True
def close(self):
if self.private_dict_file:
self.private_dict_file.close()
def _check_spelling(self, msgid, line, line_num):
original_line = line
if line.strip().startswith("#"):
line = line.strip()[1:]
starts_with_comment = True
else:
starts_with_comment = False
for word, _ in self.tokenizer(line.strip()):
lower_cased_word = word.casefold()
# Skip words from ignore list.
if word in self.ignore_list or lower_cased_word in self.ignore_list:
continue
# Strip starting u' from unicode literals and r' from raw strings.
if word.startswith(("u'", 'u"', "r'", 'r"')) and len(word) > 2:
word = word[2:]
lower_cased_word = lower_cased_word[2:]
# If it is a known word, then continue.
try:
if self.spelling_dict.check(lower_cased_word):
# The lower cased version of word passed spell checking
continue
# If we reached this far, it means there was a spelling mistake.
# Let's retry with the original work because 'unicode' is a
# spelling mistake but 'Unicode' is not
if self.spelling_dict.check(word):
continue
except enchant.errors.Error:
self.add_message(
"invalid-characters-in-docstring", line=line_num, args=(word,)
)
continue
# Store word to private dict or raise a message.
if self.config.spelling_store_unknown_words:
if lower_cased_word not in self.unknown_words:
self.private_dict_file.write("%s\n" % lower_cased_word)
self.unknown_words.add(lower_cased_word)
else:
# Present up to N suggestions.
suggestions = self.spelling_dict.suggest(word)
del suggestions[self.config.max_spelling_suggestions :]
m = re.search(r"(\W|^)(%s)(\W|$)" % word, line)
if m:
# Start position of second group in regex.
col = m.regs[2][0]
else:
col = line.index(word)
if starts_with_comment:
col += 1
indicator = (" " * col) + ("^" * len(word))
self.add_message(
msgid,
line=line_num,
args=(
word,
original_line,
indicator,
"'{}'".format("' or '".join(suggestions)),
),
)
def process_tokens(self, tokens):
if not self.initialized:
return
# Process tokens and look for comments.
for (tok_type, token, (start_row, _), _, _) in tokens:
if tok_type == tokenize.COMMENT:
if start_row == 1 and token.startswith("#!/"):
# Skip shebang lines
continue
if token.startswith("# pylint:"):
# Skip pylint enable/disable comments
continue
self._check_spelling("wrong-spelling-in-comment", token, start_row)
@check_messages("wrong-spelling-in-docstring")
def visit_module(self, node):
if not self.initialized:
return
self._check_docstring(node)
@check_messages("wrong-spelling-in-docstring")
def visit_classdef(self, node):
if not self.initialized:
return
self._check_docstring(node)
@check_messages("wrong-spelling-in-docstring")
def visit_functiondef(self, node):
if not self.initialized:
return
self._check_docstring(node)
visit_asyncfunctiondef = visit_functiondef
def _check_docstring(self, node):
"""check the node has any spelling errors"""
docstring = node.doc
if not docstring:
return
start_line = node.lineno + 1
# Go through lines of docstring
for idx, line in enumerate(docstring.splitlines()):
self._check_spelling("wrong-spelling-in-docstring", line, start_line + idx)
def register(linter):
"""required method to auto register this checker """
linter.register_checker(SpellingChecker(linter))

View File

@@ -0,0 +1,441 @@
# -*- coding: utf-8 -*-
# Copyright (c) 2013-2014 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2013-2014 Google, Inc.
# Copyright (c) 2014-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Cosmin Poieana <cmin@ropython.org>
# Copyright (c) 2014 Vlad Temian <vladtemian@gmail.com>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015 Cezar <celnazli@bitdefender.com>
# Copyright (c) 2015 Chris Rebert <code@rebertia.com>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016 Jared Garst <cultofjared@gmail.com>
# Copyright (c) 2017 Renat Galimov <renat2017@gmail.com>
# Copyright (c) 2017 Martin <MartinBasti@users.noreply.github.com>
# Copyright (c) 2017 Christopher Zurcher <zurcher@users.noreply.github.com>
# Copyright (c) 2017 Łukasz Rogalski <rogalski.91@gmail.com>
# Copyright (c) 2018 Banjamin Freeman <befreeman@users.noreply.github.com>
# Copyright (c) 2018 Ioana Tagirta <ioana.tagirta@gmail.com>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""Checkers for various standard library functions."""
import sys
import astroid
from astroid.bases import Instance
from astroid.node_classes import Const
from pylint.interfaces import IAstroidChecker
from pylint.checkers import BaseChecker
from pylint.checkers import utils
OPEN_FILES = {"open", "file"}
UNITTEST_CASE = "unittest.case"
THREADING_THREAD = "threading.Thread"
COPY_COPY = "copy.copy"
OS_ENVIRON = "os._Environ"
ENV_GETTERS = {"os.getenv"}
SUBPROCESS_POPEN = "subprocess.Popen"
if sys.version_info >= (3, 0):
OPEN_MODULE = "_io"
else:
OPEN_MODULE = "__builtin__"
def _check_mode_str(mode):
# check type
if not isinstance(mode, str):
return False
# check syntax
modes = set(mode)
_mode = "rwatb+Ux"
creating = "x" in modes
if modes - set(_mode) or len(mode) > len(modes):
return False
# check logic
reading = "r" in modes
writing = "w" in modes
appending = "a" in modes
text = "t" in modes
binary = "b" in modes
if "U" in modes:
if writing or appending or creating:
return False
reading = True
if text and binary:
return False
total = reading + writing + appending + creating
if total > 1:
return False
if not (reading or writing or appending or creating):
return False
return True
class StdlibChecker(BaseChecker):
__implements__ = (IAstroidChecker,)
name = "stdlib"
msgs = {
"W1501": (
'"%s" is not a valid mode for open.',
"bad-open-mode",
"Python supports: r, w, a[, x] modes with b, +, "
"and U (only with r) options. "
"See http://docs.python.org/2/library/functions.html#open",
),
"W1502": (
"Using datetime.time in a boolean context.",
"boolean-datetime",
"Using datetime.time in a boolean context can hide "
"subtle bugs when the time they represent matches "
"midnight UTC. This behaviour was fixed in Python 3.5. "
"See http://bugs.python.org/issue13936 for reference.",
{"maxversion": (3, 5)},
),
"W1503": (
"Redundant use of %s with constant value %r",
"redundant-unittest-assert",
"The first argument of assertTrue and assertFalse is "
"a condition. If a constant is passed as parameter, that "
"condition will be always true. In this case a warning "
"should be emitted.",
),
"W1505": (
"Using deprecated method %s()",
"deprecated-method",
"The method is marked as deprecated and will be removed in "
"a future version of Python. Consider looking for an "
"alternative in the documentation.",
),
"W1506": (
"threading.Thread needs the target function",
"bad-thread-instantiation",
"The warning is emitted when a threading.Thread class "
"is instantiated without the target function being passed. "
"By default, the first parameter is the group param, not the target param. ",
),
"W1507": (
"Using copy.copy(os.environ). Use os.environ.copy() instead. ",
"shallow-copy-environ",
"os.environ is not a dict object but proxy object, so "
"shallow copy has still effects on original object. "
"See https://bugs.python.org/issue15373 for reference. ",
),
"E1507": (
"%s does not support %s type argument",
"invalid-envvar-value",
"Env manipulation functions support only string type arguments. "
"See https://docs.python.org/3/library/os.html#os.getenv. ",
),
"W1508": (
"%s default type is %s. Expected str or None.",
"invalid-envvar-default",
"Env manipulation functions return None or str values. "
"Supplying anything different as a default may cause bugs. "
"See https://docs.python.org/3/library/os.html#os.getenv. ",
),
"W1509": (
"Using preexec_fn keyword which may be unsafe in the presence "
"of threads",
"subprocess-popen-preexec-fn",
"The preexec_fn parameter is not safe to use in the presence "
"of threads in your application. The child process could "
"deadlock before exec is called. If you must use it, keep it "
"trivial! Minimize the number of libraries you call into."
"https://docs.python.org/3/library/subprocess.html#popen-constructor",
),
}
deprecated = {
0: {
"cgi.parse_qs",
"cgi.parse_qsl",
"ctypes.c_buffer",
"distutils.command.register.register.check_metadata",
"distutils.command.sdist.sdist.check_metadata",
"tkinter.Misc.tk_menuBar",
"tkinter.Menu.tk_bindForTraversal",
},
2: {
(2, 6, 0): {
"commands.getstatus",
"os.popen2",
"os.popen3",
"os.popen4",
"macostools.touched",
},
(2, 7, 0): {
"unittest.case.TestCase.assertEquals",
"unittest.case.TestCase.assertNotEquals",
"unittest.case.TestCase.assertAlmostEquals",
"unittest.case.TestCase.assertNotAlmostEquals",
"unittest.case.TestCase.assert_",
"xml.etree.ElementTree.Element.getchildren",
"xml.etree.ElementTree.Element.getiterator",
"xml.etree.ElementTree.XMLParser.getiterator",
"xml.etree.ElementTree.XMLParser.doctype",
},
},
3: {
(3, 0, 0): {
"inspect.getargspec",
"failUnlessEqual",
"assertEquals",
"failIfEqual",
"assertNotEquals",
"failUnlessAlmostEqual",
"assertAlmostEquals",
"failIfAlmostEqual",
"assertNotAlmostEquals",
"failUnless",
"assert_",
"failUnlessRaises",
"failIf",
"assertRaisesRegexp",
"assertRegexpMatches",
"assertNotRegexpMatches",
},
(3, 1, 0): {
"base64.encodestring",
"base64.decodestring",
"ntpath.splitunc",
},
(3, 2, 0): {
"cgi.escape",
"configparser.RawConfigParser.readfp",
"xml.etree.ElementTree.Element.getchildren",
"xml.etree.ElementTree.Element.getiterator",
"xml.etree.ElementTree.XMLParser.getiterator",
"xml.etree.ElementTree.XMLParser.doctype",
},
(3, 3, 0): {
"inspect.getmoduleinfo",
"logging.warn",
"logging.Logger.warn",
"logging.LoggerAdapter.warn",
"nntplib._NNTPBase.xpath",
"platform.popen",
},
(3, 4, 0): {
"importlib.find_loader",
"plistlib.readPlist",
"plistlib.writePlist",
"plistlib.readPlistFromBytes",
"plistlib.writePlistToBytes",
},
(3, 4, 4): {"asyncio.tasks.async"},
(3, 5, 0): {
"fractions.gcd",
"inspect.getargvalues",
"inspect.formatargspec",
"inspect.formatargvalues",
"inspect.getcallargs",
"platform.linux_distribution",
"platform.dist",
},
(3, 6, 0): {"importlib._bootstrap_external.FileLoader.load_module"},
},
}
def _check_bad_thread_instantiation(self, node):
if not node.kwargs and not node.keywords and len(node.args) <= 1:
self.add_message("bad-thread-instantiation", node=node)
def _check_for_preexec_fn_in_Popen(self, node):
if node.keywords:
for keyword in node.keywords:
if keyword.arg == "preexec_fn":
self.add_message("subprocess-popen-preexec-fn", node=node)
def _check_shallow_copy_environ(self, node):
arg = utils.get_argument_from_call(node, position=0)
for inferred in arg.inferred():
if inferred.qname() == OS_ENVIRON:
self.add_message("shallow-copy-environ", node=node)
break
@utils.check_messages(
"bad-open-mode",
"redundant-unittest-assert",
"deprecated-method",
"bad-thread-instantiation",
"shallow-copy-environ",
"invalid-envvar-value",
"invalid-envvar-default",
"subprocess-popen-preexec-fn",
)
def visit_call(self, node):
"""Visit a Call node."""
try:
for inferred in node.func.infer():
if inferred is astroid.Uninferable:
continue
elif inferred.root().name == OPEN_MODULE:
if getattr(node.func, "name", None) in OPEN_FILES:
self._check_open_mode(node)
elif inferred.root().name == UNITTEST_CASE:
self._check_redundant_assert(node, inferred)
elif isinstance(inferred, astroid.ClassDef):
if inferred.qname() == THREADING_THREAD:
self._check_bad_thread_instantiation(node)
elif inferred.qname() == SUBPROCESS_POPEN:
self._check_for_preexec_fn_in_Popen(node)
elif isinstance(inferred, astroid.FunctionDef):
name = inferred.qname()
if name == COPY_COPY:
self._check_shallow_copy_environ(node)
elif name in ENV_GETTERS:
self._check_env_function(node, inferred)
self._check_deprecated_method(node, inferred)
except astroid.InferenceError:
return
@utils.check_messages("boolean-datetime")
def visit_unaryop(self, node):
if node.op == "not":
self._check_datetime(node.operand)
@utils.check_messages("boolean-datetime")
def visit_if(self, node):
self._check_datetime(node.test)
@utils.check_messages("boolean-datetime")
def visit_ifexp(self, node):
self._check_datetime(node.test)
@utils.check_messages("boolean-datetime")
def visit_boolop(self, node):
for value in node.values:
self._check_datetime(value)
def _check_deprecated_method(self, node, inferred):
py_vers = sys.version_info[0]
if isinstance(node.func, astroid.Attribute):
func_name = node.func.attrname
elif isinstance(node.func, astroid.Name):
func_name = node.func.name
else:
# Not interested in other nodes.
return
# Reject nodes which aren't of interest to us.
acceptable_nodes = (
astroid.BoundMethod,
astroid.UnboundMethod,
astroid.FunctionDef,
)
if not isinstance(inferred, acceptable_nodes):
return
qname = inferred.qname()
if any(name in self.deprecated[0] for name in (qname, func_name)):
self.add_message("deprecated-method", node=node, args=(func_name,))
else:
for since_vers, func_list in self.deprecated[py_vers].items():
if since_vers <= sys.version_info and any(
name in func_list for name in (qname, func_name)
):
self.add_message("deprecated-method", node=node, args=(func_name,))
break
def _check_redundant_assert(self, node, infer):
if (
isinstance(infer, astroid.BoundMethod)
and node.args
and isinstance(node.args[0], astroid.Const)
and infer.name in ["assertTrue", "assertFalse"]
):
self.add_message(
"redundant-unittest-assert",
args=(infer.name, node.args[0].value),
node=node,
)
def _check_datetime(self, node):
""" Check that a datetime was infered.
If so, emit boolean-datetime warning.
"""
try:
infered = next(node.infer())
except astroid.InferenceError:
return
if isinstance(infered, Instance) and infered.qname() == "datetime.time":
self.add_message("boolean-datetime", node=node)
def _check_open_mode(self, node):
"""Check that the mode argument of an open or file call is valid."""
try:
mode_arg = utils.get_argument_from_call(node, position=1, keyword="mode")
except utils.NoSuchArgumentError:
return
if mode_arg:
mode_arg = utils.safe_infer(mode_arg)
if isinstance(mode_arg, astroid.Const) and not _check_mode_str(
mode_arg.value
):
self.add_message("bad-open-mode", node=node, args=mode_arg.value)
def _check_env_function(self, node, infer):
env_name_kwarg = "key"
env_value_kwarg = "default"
if node.keywords:
kwargs = {keyword.arg: keyword.value for keyword in node.keywords}
else:
kwargs = None
if node.args:
env_name_arg = node.args[0]
elif kwargs and env_name_kwarg in kwargs:
env_name_arg = kwargs[env_name_kwarg]
else:
env_name_arg = None
if env_name_arg:
self._check_invalid_envvar_value(
node=node,
message="invalid-envvar-value",
call_arg=utils.safe_infer(env_name_arg),
infer=infer,
allow_none=False,
)
if len(node.args) == 2:
env_value_arg = node.args[1]
elif kwargs and env_value_kwarg in kwargs:
env_value_arg = kwargs[env_value_kwarg]
else:
env_value_arg = None
if env_value_arg:
self._check_invalid_envvar_value(
node=node,
infer=infer,
message="invalid-envvar-default",
call_arg=utils.safe_infer(env_value_arg),
allow_none=True,
)
def _check_invalid_envvar_value(self, node, infer, message, call_arg, allow_none):
if call_arg in (astroid.Uninferable, None):
return
name = infer.qname()
if isinstance(call_arg, Const):
emit = False
if call_arg.value is None:
emit = not allow_none
elif not isinstance(call_arg.value, str):
emit = True
if emit:
self.add_message(message, node=node, args=(name, call_arg.pytype()))
else:
self.add_message(message, node=node, args=(name, call_arg.pytype()))
def register(linter):
"""required method to auto register this checker """
linter.register_checker(StdlibChecker(linter))

View File

@@ -0,0 +1,763 @@
# -*- coding: utf-8 -*-
# Copyright (c) 2009 Charles Hebert <charles.hebert@logilab.fr>
# Copyright (c) 2010-2014 LOGILAB S.A. (Paris, FRANCE) <contact@logilab.fr>
# Copyright (c) 2010 Daniel Harding <dharding@gmail.com>
# Copyright (c) 2012-2014 Google, Inc.
# Copyright (c) 2013-2018 Claudiu Popa <pcmanticore@gmail.com>
# Copyright (c) 2014 Brett Cannon <brett@python.org>
# Copyright (c) 2014 Arun Persaud <arun@nubati.net>
# Copyright (c) 2015 Rene Zhang <rz99@cornell.edu>
# Copyright (c) 2015 Ionel Cristian Maries <contact@ionelmc.ro>
# Copyright (c) 2016, 2018 Jakub Wilk <jwilk@jwilk.net>
# Copyright (c) 2016 Peter Dawyndt <Peter.Dawyndt@UGent.be>
# Copyright (c) 2017 Łukasz Rogalski <rogalski.91@gmail.com>
# Copyright (c) 2017 Ville Skyttä <ville.skytta@iki.fi>
# Copyright (c) 2018 Nick Drozd <nicholasdrozd@gmail.com>
# Copyright (c) 2018 Anthony Sottile <asottile@umich.edu>
# Licensed under the GPL: https://www.gnu.org/licenses/old-licenses/gpl-2.0.html
# For details: https://github.com/PyCQA/pylint/blob/master/COPYING
"""Checker for string formatting operations.
"""
import builtins
import sys
import tokenize
import numbers
from collections import Counter
import astroid
from astroid.arguments import CallSite
from astroid.node_classes import Const
from pylint.interfaces import ITokenChecker, IAstroidChecker, IRawChecker
from pylint.checkers import BaseChecker, BaseTokenChecker
from pylint.checkers import utils
from pylint.checkers.utils import check_messages
_AST_NODE_STR_TYPES = ("__builtin__.unicode", "__builtin__.str", "builtins.str")
_PY3K = sys.version_info[:2] >= (3, 0)
_PY27 = sys.version_info[:2] == (2, 7)
MSGS = {
"E1300": (
"Unsupported format character %r (%#02x) at index %d",
"bad-format-character",
"Used when an unsupported format character is used in a format string.",
),
"E1301": (
"Format string ends in middle of conversion specifier",
"truncated-format-string",
"Used when a format string terminates before the end of a "
"conversion specifier.",
),
"E1302": (
"Mixing named and unnamed conversion specifiers in format string",
"mixed-format-string",
"Used when a format string contains both named (e.g. '%(foo)d') "
"and unnamed (e.g. '%d') conversion specifiers. This is also "
"used when a named conversion specifier contains * for the "
"minimum field width and/or precision.",
),
"E1303": (
"Expected mapping for format string, not %s",
"format-needs-mapping",
"Used when a format string that uses named conversion specifiers "
"is used with an argument that is not a mapping.",
),
"W1300": (
"Format string dictionary key should be a string, not %s",
"bad-format-string-key",
"Used when a format string that uses named conversion specifiers "
"is used with a dictionary whose keys are not all strings.",
),
"W1301": (
"Unused key %r in format string dictionary",
"unused-format-string-key",
"Used when a format string that uses named conversion specifiers "
"is used with a dictionary that contains keys not required by the "
"format string.",
),
"E1304": (
"Missing key %r in format string dictionary",
"missing-format-string-key",
"Used when a format string that uses named conversion specifiers "
"is used with a dictionary that doesn't contain all the keys "
"required by the format string.",
),
"E1305": (
"Too many arguments for format string",
"too-many-format-args",
"Used when a format string that uses unnamed conversion "
"specifiers is given too many arguments.",
),
"E1306": (
"Not enough arguments for format string",
"too-few-format-args",
"Used when a format string that uses unnamed conversion "
"specifiers is given too few arguments",
),
"E1307": (
"Argument %r does not match format type %r",
"bad-string-format-type",
"Used when a type required by format string "
"is not suitable for actual argument type",
),
"E1310": (
"Suspicious argument in %s.%s call",
"bad-str-strip-call",
"The argument to a str.{l,r,}strip call contains a duplicate character, ",
),
"W1302": (
"Invalid format string",
"bad-format-string",
"Used when a PEP 3101 format string is invalid.",
),
"W1303": (
"Missing keyword argument %r for format string",
"missing-format-argument-key",
"Used when a PEP 3101 format string that uses named fields "
"doesn't receive one or more required keywords.",
),
"W1304": (
"Unused format argument %r",
"unused-format-string-argument",
"Used when a PEP 3101 format string that uses named "
"fields is used with an argument that "
"is not required by the format string.",
),
"W1305": (
"Format string contains both automatic field numbering "
"and manual field specification",
"format-combined-specification",
"Used when a PEP 3101 format string contains both automatic "
"field numbering (e.g. '{}') and manual field "
"specification (e.g. '{0}').",
),
"W1306": (
"Missing format attribute %r in format specifier %r",
"missing-format-attribute",
"Used when a PEP 3101 format string uses an "
"attribute specifier ({0.length}), but the argument "
"passed for formatting doesn't have that attribute.",
),
"W1307": (
"Using invalid lookup key %r in format specifier %r",
"invalid-format-index",
"Used when a PEP 3101 format string uses a lookup specifier "
"({a[1]}), but the argument passed for formatting "
"doesn't contain or doesn't have that key as an attribute.",
),
"W1308": (
"Duplicate string formatting argument %r, consider passing as named argument",
"duplicate-string-formatting-argument",
"Used when we detect that a string formatting is "
"repeating an argument instead of using named string arguments",
),
}
OTHER_NODES = (
astroid.Const,
astroid.List,
astroid.Lambda,
astroid.FunctionDef,
astroid.ListComp,
astroid.SetComp,
astroid.GeneratorExp,
)
BUILTINS_STR = builtins.__name__ + ".str"
BUILTINS_FLOAT = builtins.__name__ + ".float"
BUILTINS_INT = builtins.__name__ + ".int"
def get_access_path(key, parts):
""" Given a list of format specifiers, returns
the final access path (e.g. a.b.c[0][1]).
"""
path = []
for is_attribute, specifier in parts:
if is_attribute:
path.append(".{}".format(specifier))
else:
path.append("[{!r}]".format(specifier))
return str(key) + "".join(path)
def arg_matches_format_type(arg_type, format_type):
if format_type in "sr":
# All types can be printed with %s and %r
return True
if isinstance(arg_type, astroid.Instance):
arg_type = arg_type.pytype()
if arg_type == BUILTINS_STR:
return format_type == "c"
if arg_type == BUILTINS_FLOAT:
return format_type in "deEfFgGn%"
if arg_type == BUILTINS_INT:
# Integers allow all types
return True
return False
return True
class StringFormatChecker(BaseChecker):
"""Checks string formatting operations to ensure that the format string
is valid and the arguments match the format string.
"""
__implements__ = (IAstroidChecker,)
name = "string"
msgs = MSGS
# pylint: disable=too-many-branches
@check_messages(*MSGS)
def visit_binop(self, node):
if node.op != "%":
return
left = node.left
args = node.right
if not (isinstance(left, astroid.Const) and isinstance(left.value, str)):
return
format_string = left.value
try:
required_keys, required_num_args, required_key_types, required_arg_types = utils.parse_format_string(
format_string
)
except utils.UnsupportedFormatCharacter as e:
c = format_string[e.index]
self.add_message(
"bad-format-character", node=node, args=(c, ord(c), e.index)
)
return
except utils.IncompleteFormatString:
self.add_message("truncated-format-string", node=node)
return
if required_keys and required_num_args:
# The format string uses both named and unnamed format
# specifiers.
self.add_message("mixed-format-string", node=node)
elif required_keys:
# The format string uses only named format specifiers.
# Check that the RHS of the % operator is a mapping object
# that contains precisely the set of keys required by the
# format string.
if isinstance(args, astroid.Dict):
keys = set()
unknown_keys = False
for k, _ in args.items:
if isinstance(k, astroid.Const):
key = k.value
if isinstance(key, str):
keys.add(key)
else:
self.add_message(
"bad-format-string-key", node=node, args=key
)
else:
# One of the keys was something other than a
# constant. Since we can't tell what it is,
# suppress checks for missing keys in the
# dictionary.
unknown_keys = True
if not unknown_keys:
for key in required_keys:
if key not in keys:
self.add_message(
"missing-format-string-key", node=node, args=key
)
for key in keys:
if key not in required_keys:
self.add_message(
"unused-format-string-key", node=node, args=key
)
for key, arg in args.items:
if not isinstance(key, astroid.Const):
continue
format_type = required_key_types.get(key.value, None)
arg_type = utils.safe_infer(arg)
if (
format_type is not None
and arg_type not in (None, astroid.Uninferable)
and not arg_matches_format_type(arg_type, format_type)
):
self.add_message(
"bad-string-format-type",
node=node,
args=(arg_type.pytype(), format_type),
)
elif isinstance(args, (OTHER_NODES, astroid.Tuple)):
type_name = type(args).__name__
self.add_message("format-needs-mapping", node=node, args=type_name)
# else:
# The RHS of the format specifier is a name or
# expression. It may be a mapping object, so
# there's nothing we can check.
else:
# The format string uses only unnamed format specifiers.
# Check that the number of arguments passed to the RHS of
# the % operator matches the number required by the format
# string.
args_elts = ()
if isinstance(args, astroid.Tuple):
rhs_tuple = utils.safe_infer(args)
num_args = None
if hasattr(rhs_tuple, "elts"):
args_elts = rhs_tuple.elts
num_args = len(args_elts)
elif isinstance(args, (OTHER_NODES, (astroid.Dict, astroid.DictComp))):
args_elts = [args]
num_args = 1
else:
# The RHS of the format specifier is a name or
# expression. It could be a tuple of unknown size, so
# there's nothing we can check.
num_args = None
if num_args is not None:
if num_args > required_num_args:
self.add_message("too-many-format-args", node=node)
elif num_args < required_num_args:
self.add_message("too-few-format-args", node=node)
for arg, format_type in zip(args_elts, required_arg_types):
if not arg:
continue
arg_type = utils.safe_infer(arg)
if arg_type not in (
None,
astroid.Uninferable,
) and not arg_matches_format_type(arg_type, format_type):
self.add_message(
"bad-string-format-type",
node=node,
args=(arg_type.pytype(), format_type),
)
@check_messages(*MSGS)
def visit_call(self, node):
func = utils.safe_infer(node.func)
if (
isinstance(func, astroid.BoundMethod)
and isinstance(func.bound, astroid.Instance)
and func.bound.name in ("str", "unicode", "bytes")
):
if func.name in ("strip", "lstrip", "rstrip") and node.args:
arg = utils.safe_infer(node.args[0])
if not isinstance(arg, astroid.Const) or not isinstance(arg.value, str):
return
if len(arg.value) != len(set(arg.value)):
self.add_message(
"bad-str-strip-call",
node=node,
args=(func.bound.name, func.name),
)
elif func.name == "format":
self._check_new_format(node, func)
def _detect_vacuous_formatting(self, node, positional_arguments):
counter = Counter(
arg.name for arg in positional_arguments if isinstance(arg, astroid.Name)
)
for name, count in counter.items():
if count == 1:
continue
self.add_message(
"duplicate-string-formatting-argument", node=node, args=(name,)
)
def _check_new_format(self, node, func):
""" Check the new string formatting. """
# TODO: skip (for now) format nodes which don't have
# an explicit string on the left side of the format operation.
# We do this because our inference engine can't properly handle
# redefinitions of the original string.
# For more details, see issue 287.
#
# Note that there may not be any left side at all, if the format method
# has been assigned to another variable. See issue 351. For example:
#
# fmt = 'some string {}'.format
# fmt('arg')
if isinstance(node.func, astroid.Attribute) and not isinstance(
node.func.expr, astroid.Const
):
return
if node.starargs or node.kwargs:
return
try:
strnode = next(func.bound.infer())
except astroid.InferenceError:
return
if not (isinstance(strnode, astroid.Const) and isinstance(strnode.value, str)):
return
try:
call_site = CallSite.from_call(node)
except astroid.InferenceError:
return
try:
fields, num_args, manual_pos = utils.parse_format_method_string(
strnode.value
)
except utils.IncompleteFormatString:
self.add_message("bad-format-string", node=node)
return
positional_arguments = call_site.positional_arguments
named_arguments = call_site.keyword_arguments
named_fields = {field[0] for field in fields if isinstance(field[0], str)}
if num_args and manual_pos:
self.add_message("format-combined-specification", node=node)
return
check_args = False
# Consider "{[0]} {[1]}" as num_args.
num_args += sum(1 for field in named_fields if field == "")
if named_fields:
for field in named_fields:
if field and field not in named_arguments:
self.add_message(
"missing-format-argument-key", node=node, args=(field,)
)
for field in named_arguments:
if field not in named_fields:
self.add_message(
"unused-format-string-argument", node=node, args=(field,)
)
# num_args can be 0 if manual_pos is not.
num_args = num_args or manual_pos
if positional_arguments or num_args:
empty = any(True for field in named_fields if field == "")
if named_arguments or empty:
# Verify the required number of positional arguments
# only if the .format got at least one keyword argument.
# This means that the format strings accepts both
# positional and named fields and we should warn
# when one of the them is missing or is extra.
check_args = True
else:
check_args = True
if check_args:
# num_args can be 0 if manual_pos is not.
num_args = num_args or manual_pos
if len(positional_arguments) > num_args:
self.add_message("too-many-format-args", node=node)
elif len(positional_arguments) < num_args:
self.add_message("too-few-format-args", node=node)
self._detect_vacuous_formatting(node, positional_arguments)
self._check_new_format_specifiers(node, fields, named_arguments)
def _check_new_format_specifiers(self, node, fields, named):
"""
Check attribute and index access in the format
string ("{0.a}" and "{0[a]}").
"""
for key, specifiers in fields:
# Obtain the argument. If it can't be obtained
# or infered, skip this check.
if key == "":
# {[0]} will have an unnamed argument, defaulting
# to 0. It will not be present in `named`, so use the value
# 0 for it.
key = 0
if isinstance(key, numbers.Number):
try:
argname = utils.get_argument_from_call(node, key)
except utils.NoSuchArgumentError:
continue
else:
if key not in named:
continue
argname = named[key]
if argname in (astroid.Uninferable, None):
continue
try:
argument = next(argname.infer())
except astroid.InferenceError:
continue
if not specifiers or argument is astroid.Uninferable:
# No need to check this key if it doesn't
# use attribute / item access
continue
if argument.parent and isinstance(argument.parent, astroid.Arguments):
# Ignore any object coming from an argument,
# because we can't infer its value properly.
continue
previous = argument
parsed = []
for is_attribute, specifier in specifiers:
if previous is astroid.Uninferable:
break
parsed.append((is_attribute, specifier))
if is_attribute:
try:
previous = previous.getattr(specifier)[0]
except astroid.NotFoundError:
if (
hasattr(previous, "has_dynamic_getattr")
and previous.has_dynamic_getattr()
):
# Don't warn if the object has a custom __getattr__
break
path = get_access_path(key, parsed)
self.add_message(
"missing-format-attribute",
args=(specifier, path),
node=node,
)
break
else:
warn_error = False
if hasattr(previous, "getitem"):
try:
previous = previous.getitem(astroid.Const(specifier))
except (
astroid.AstroidIndexError,
astroid.AstroidTypeError,
astroid.AttributeInferenceError,
):
warn_error = True
except astroid.InferenceError:
break
if previous is astroid.Uninferable:
break
else:
try:
# Lookup __getitem__ in the current node,
# but skip further checks, because we can't
# retrieve the looked object
previous.getattr("__getitem__")
break
except astroid.NotFoundError:
warn_error = True
if warn_error:
path = get_access_path(key, parsed)
self.add_message(
"invalid-format-index", args=(specifier, path), node=node
)
break
try:
previous = next(previous.infer())
except astroid.InferenceError:
# can't check further if we can't infer it
break
class StringConstantChecker(BaseTokenChecker):
"""Check string literals"""
__implements__ = (IAstroidChecker, ITokenChecker, IRawChecker)
name = "string"
msgs = {
"W1401": (
"Anomalous backslash in string: '%s'. "
"String constant might be missing an r prefix.",
"anomalous-backslash-in-string",
"Used when a backslash is in a literal string but not as an escape.",
),
"W1402": (
"Anomalous Unicode escape in byte string: '%s'. "
"String constant might be missing an r or u prefix.",
"anomalous-unicode-escape-in-string",
"Used when an escape like \\u is encountered in a byte "
"string where it has no effect.",
),
"W1403": (
"Implicit string concatenation found in %s",
"implicit-str-concat-in-sequence",
"String literals are implicitly concatenated in a "
"literal iterable definition : "
"maybe a comma is missing ?",
),
}
options = (
(
"check-str-concat-over-line-jumps",
{
"default": False,
"type": "yn",
"metavar": "<y_or_n>",
"help": "This flag controls whether the "
"implicit-str-concat-in-sequence should generate a warning "
"on implicit string concatenation in sequences defined over "
"several lines.",
},
),
)
# Characters that have a special meaning after a backslash in either
# Unicode or byte strings.
ESCAPE_CHARACTERS = "abfnrtvx\n\r\t\\'\"01234567"
# TODO(mbp): Octal characters are quite an edge case today; people may
# prefer a separate warning where they occur. \0 should be allowed.
# Characters that have a special meaning after a backslash but only in
# Unicode strings.
UNICODE_ESCAPE_CHARACTERS = "uUN"
def __init__(self, *args, **kwargs):
super(StringConstantChecker, self).__init__(*args, **kwargs)
self.string_tokens = {} # token position -> (token value, next token)
def process_module(self, module):
self._unicode_literals = "unicode_literals" in module.future_imports
def process_tokens(self, tokens):
encoding = "ascii"
for i, (tok_type, token, start, _, line) in enumerate(tokens):
if tok_type == tokenize.ENCODING:
# this is always the first token processed
encoding = token
elif tok_type == tokenize.STRING:
# 'token' is the whole un-parsed token; we can look at the start
# of it to see whether it's a raw or unicode string etc.
self.process_string_token(token, start[0])
# We figure the next token, ignoring comments & newlines:
j = i + 1
while j < len(tokens) and tokens[j].type in (
tokenize.NEWLINE,
tokenize.NL,
tokenize.COMMENT,
):
j += 1
next_token = tokens[j] if j < len(tokens) else None
if encoding != "ascii":
# We convert `tokenize` character count into a byte count,
# to match with astroid `.col_offset`
start = (start[0], len(line[: start[1]].encode(encoding)))
self.string_tokens[start] = (str_eval(token), next_token)
@check_messages(*(msgs.keys()))
def visit_list(self, node):
self.check_for_concatenated_strings(node, "list")
@check_messages(*(msgs.keys()))
def visit_set(self, node):
self.check_for_concatenated_strings(node, "set")
@check_messages(*(msgs.keys()))
def visit_tuple(self, node):
self.check_for_concatenated_strings(node, "tuple")
def check_for_concatenated_strings(self, iterable_node, iterable_type):
for elt in iterable_node.elts:
if isinstance(elt, Const) and elt.pytype() in _AST_NODE_STR_TYPES:
if elt.col_offset < 0:
# This can happen in case of escaped newlines
continue
if (elt.lineno, elt.col_offset) not in self.string_tokens:
# This may happen with Latin1 encoding
# cf. https://github.com/PyCQA/pylint/issues/2610
continue
matching_token, next_token = self.string_tokens[
(elt.lineno, elt.col_offset)
]
# We detect string concatenation: the AST Const is the
# combination of 2 string tokens
if matching_token != elt.value and next_token is not None:
if next_token.type == tokenize.STRING and (
next_token.start[0] == elt.lineno
or self.config.check_str_concat_over_line_jumps
):
self.add_message(
"implicit-str-concat-in-sequence",
line=elt.lineno,
args=(iterable_type,),
)
def process_string_token(self, token, start_row):
quote_char = None
index = None
for index, c in enumerate(token):
if c in "'\"":
quote_char = c
break
if quote_char is None:
return
prefix = token[:index].lower() # markers like u, b, r.
after_prefix = token[index:]
if after_prefix[:3] == after_prefix[-3:] == 3 * quote_char:
string_body = after_prefix[3:-3]
else:
string_body = after_prefix[1:-1] # Chop off quotes
# No special checks on raw strings at the moment.
if "r" not in prefix:
self.process_non_raw_string_token(prefix, string_body, start_row)
def process_non_raw_string_token(self, prefix, string_body, start_row):
"""check for bad escapes in a non-raw string.
prefix: lowercase string of eg 'ur' string prefix markers.
string_body: the un-parsed body of the string, not including the quote
marks.
start_row: integer line number in the source.
"""
# Walk through the string; if we see a backslash then escape the next
# character, and skip over it. If we see a non-escaped character,
# alert, and continue.
#
# Accept a backslash when it escapes a backslash, or a quote, or
# end-of-line, or one of the letters that introduce a special escape
# sequence <http://docs.python.org/reference/lexical_analysis.html>
#
# TODO(mbp): Maybe give a separate warning about the rarely-used
# \a \b \v \f?
#
# TODO(mbp): We could give the column of the problem character, but
# add_message doesn't seem to have a way to pass it through at present.
i = 0
while True:
i = string_body.find("\\", i)
if i == -1:
break
# There must be a next character; having a backslash at the end
# of the string would be a SyntaxError.
next_char = string_body[i + 1]
match = string_body[i : i + 2]
if next_char in self.UNICODE_ESCAPE_CHARACTERS:
if "u" in prefix:
pass
elif (_PY3K or self._unicode_literals) and "b" not in prefix:
pass # unicode by default
else:
self.add_message(
"anomalous-unicode-escape-in-string",
line=start_row,
args=(match,),
)
elif next_char not in self.ESCAPE_CHARACTERS:
self.add_message(
"anomalous-backslash-in-string", line=start_row, args=(match,)
)
# Whether it was a valid escape or not, backslash followed by
# another character can always be consumed whole: the second
# character can never be the start of a new backslash escape.
i += 2
def register(linter):
"""required method to auto register this checker """
linter.register_checker(StringFormatChecker(linter))
linter.register_checker(StringConstantChecker(linter))
def str_eval(token):
"""
Mostly replicate `ast.literal_eval(token)` manually to avoid any performance hit.
This supports f-strings, contrary to `ast.literal_eval`.
We have to support all string literal notations:
https://docs.python.org/3/reference/lexical_analysis.html#string-and-bytes-literals
"""
if token[0:2].lower() in ("fr", "rf"):
token = token[2:]
elif token[0].lower() in ("r", "u", "f"):
token = token[1:]
if token[0:3] in ('"""', "'''"):
return token[3:-3]
return token[1:-1]

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff